hana_sql_ref_en

SAP HANA Database - SQL Reference ManualSAP HANA Appliance Software SPS 05, Revision 45

Target Audience

ConsultantsAdministratorsSAP Hardware PartnerOthers

1 2012-12-19

Copyright © Copyright 2012 SAP AG. All rights reserved

SAP Library document classification: PUBLIC

No part of this publication may be reproduced or transmitted in any form or forany purpose without the express permission of SAP AG. The informationcontained herein may be changed without prior notice.

Some software products marketed by SAP AG and its distributors containproprietary software components of other software vendors.

Microsoft, Windows, Excel, Outlook, PowerPoint, Silverlight, and Visual Studioare registered trademarks of Microsoft Corporation.

IBM, DB2, DB2 Universal Database, System i, System i5, System p, Systemp5, System x, System z, System z10, z10, z/VM, z/OS, OS/390, zEnterprise,PowerVM, Power Architecture, Power Systems, POWER7, POWER6+,POWER6, POWER, PowerHA, pureScale, PowerPC, BladeCenter, SystemStorage, Storwize, XIV, GPFS, HACMP, RETAIN, DB2 Connect, RACF,Redbooks, OS/2, AIX, Intelligent Miner, WebSphere, Tivoli, Informix, andSmarter Planet are trademarks or registered trademarks of IBM Corporation.

Linux is the registered trademark of Linus Torvalds in the United States andother countries.

Adobe, the Adobe logo, Acrobat, PostScript, and Reader are trademarks orregistered trademarks of Adobe Systems Incorporated in the United States andother countries.

Oracle and Java are registered trademarks of Oracle and its affiliates.

UNIX, X/Open, OSF/1, and Motif are registered trademarks of the Open Group.

Citrix, ICA, Program Neighborhood, MetaFrame, WinFrame, VideoFrame, andMultiWin are trademarks or registered trademarks of Citrix Systems Inc.

HTML, XML, XHTML, and W3C are trademarks or registered trademarks ofW3C®, World Wide Web Consortium, Massachusetts Institute of Technology.

Apple, App Store, iBooks, iPad, iPhone, iPhoto, iPod, iTunes, Multi-Touch,Objective-C, Retina, Safari, Siri, and Xcode are trademarks or registeredtrademarks of Apple Inc.

IOS is a registered trademark of Cisco Systems Inc.

RIM, BlackBerry, BBM, BlackBerry Curve, BlackBerry Bold, BlackBerry Pearl,

2 2012-12-19

BlackBerry Torch, BlackBerry Storm, BlackBerry Storm2, BlackBerry PlayBook,and BlackBerry App World are trademarks or registered trademarks of Researchin Motion Limited.

Google App Engine, Google Apps, Google Checkout, Google Data API, GoogleMaps, Google Mobile Ads, Google Mobile Updater, Google Mobile, Google Store,Google Sync, Google Updater, Google Voice, Google Mail, Gmail, YouTube,Dalvik and Android are trademarks or registered trademarks of Google Inc.

INTERMEC is a registered trademark of Intermec Technologies Corporation.

Wi-Fi is a registered trademark of Wi-Fi Alliance.

Bluetooth is a registered trademark of Bluetooth SIG Inc.

Motorola is a registered trademark of Motorola Trademark Holdings LLC.

Computop is a registered trademark of Computop Wirtschaftsinformatik GmbH.

SAP, R/3, SAP NetWeaver, Duet, PartnerEdge, ByDesign, SAPBusinessObjects Explorer, StreamWork, SAP HANA, and other SAP productsand services mentioned herein as well as their respective logos are trademarksor registered trademarks of SAP AG in Germany and other countries.

Business Objects and the Business Objects logo, BusinessObjects, CrystalReports, Crystal Decisions, Web Intelligence, Xcelsius, and other BusinessObjects products and services mentioned herein as well as their respectivelogos are trademarks or registered trademarks of Business Objects SoftwareLtd. Business Objects is an SAP company.

Sybase and Adaptive Server, iAnywhere, Sybase 365, SQL Anywhere, and otherSybase products and services mentioned herein as well as their respective logosare trademarks or registered trademarks of Sybase Inc. Sybase is an SAPcompany.

Crossgate, m@gic EDDY, B2B 360°, and B2B 360° Services are registeredtrademarks of Crossgate AG in Germany and other countries. Crossgate is anSAP company.

All other product and service names mentioned are the trademarks of theirrespective companies. Data contained in this document serves informationalpurposes only. National product specifications may vary.

These materials are subject to change without notice. These materials areprovided by SAP AG and its affiliated companies ("SAP Group") for informationalpurposes only, without representation or warranty of any kind, and SAP Groupshall not be liable for errors or omissions with respect to the materials. The onlywarranties for SAP Group products and services are those that are set forth inthe express warranty statements accompanying such products and services, ifany. Nothing herein should be construed as constituting an additional warranty.

3 2012-12-19

489

101010

1111111111111112

1313131314141415

15161717181820

2222222222222223

2525252526262626

2828282828

303030303030313131

Table of Content

Table of ContentSAP HANA Database - SQL Reference Manual

SQL Reference ManualNotation

Symbols used in BNFBNF Lowest Terms Representations

IntroductionSQLSupported Languages and Code PagesCommentIdentifiersSingle Quotation MarkDouble Quotation MarkSQL Reserved Words

Data TypesClassification of Data TypesDatetime Types

Date FormatsTime FormatsTimestamp FormatsAdditional FormatsSupported Functions for Date/Time types

Numeric TypesCharacter String TypesBinary TypesLarge Object (LOB) TypesMapping between SQL Data Type and Column Store Data TypeData Type ConversionTyped Constant

PredicatesComparison PredicatesRange PredicateIn PredicateExists PredicateLIKE PredicateNULL PredicateCONTAINS Predicate

OperatorsUnary and Binary OperatorsOperator PrecedenceArithmetic OperatorsString OperatorsComparsion OperatorsLogical OperatorsSet Operators

Express ionsCase ExpressionsFunction ExpressionsAggregate ExpressionsSubqueries in Expressions

SQL FunctionsIntroductionData Type Conversion Functions

CASTTO_ALPHANUMTO_BIGINTTO_BINARYTO_BLOBTO_CLOB

Table of Content

4 2012-12-19

31323232333333333434343435353536

3636363637373737383838383939393940404040414141424242424343434344

44444444454545454646464747474748484848494949505050505151

TO_DATETO_DATSTO_DECIMALTO_DOUBLETO_INTTO_INTEGERTO_NCLOBTO_NVARCHARTO_REALTO_SECONDDATETO_SMALLDECIMALTO_SMALLINTTO_TIMETO_TIMESTAMPTO_TINYINTTO_VARCHAR

DateTime FunctionsADD_DAYSADD_MONTHSADD_SECONDSADD_YEARSCURRENT_DATECURRENT_TIMECURRENT_TIMESTAMPCURRENT_UTCDATECURRENT_UTCTIMECURRENT_UTCTIMESTAMPDAYNAMEDAYOFMONTHDAYOFYEARDAYS_BETWEENEXTRACTHOURISOWEEKLAST_DAYLOCALTOUTCMINUTEMONTHMONTHNAMENEXT_DAYNOWQUARTERSECONDSECONDS_BETWEENUTCTOLOCALWEEKWEEKDAYYEAR

Number FunctionsABSACOSASINATANATAN2BINTOHEXBITANDCEILCOSCOSHCOTEXPFLOORGREATESTHEXTOBINLEASTLNLOGMODPOWERRANDROUNDSIGNSINSINHSQRTTAN

Table of Content

5 2012-12-19

5151

5252525253535353545454555555565656565757575858

58585859595960616162626263

646464656667687781828384858695

101102102103103104104105106106107108108

109109110114114115117118128

TANHUMINUS

String FunctionsASCIICHARCONCATLCASELEFTLENGTHLOCATELOWERLPADLTRIMNCHARREPLACERIGHTRPADRTRIMSUBSTR_AFTERSUBSTR_BEFORESUBSTRINGTRIMUCASEUNICODEUPPER

Window FunctionsMiscellaneous Functions

COALESCECURRENT_CONNECTIONCURRENT_SCHEMACURRENT_USERGROUPING_IDIFNULLMAPNULLIFSESSION_CONTEXTSESSION_USERSYSUUID

SQL StatementsData Definition Statements

ALTER AUDIT POLICYALTER FULLTEXT INDEXALTER INDEXALTER SEQUENCEALTER TABLECREATE AUDIT POLICYCREATE FULLTEXT INDEXCREATE INDEXCREATE SCHEMACREATE SEQUENCECREATE SYNONYMCREATE TABLECREATE TRIGGERCREATE VIEWDROP AUDIT POLICYDROP FULLTEXT INDEXDROP INDEXDROP SCHEMADROP SEQUENCEDROP SYNONYMDROP TABLEDROP TRIGGERDROP VIEWRENAME COLUMNRENAME INDEXRENAME TABLE

Data Manipulation StatementsDELETEEXPLAIN PLANINSERTLOADMERGE DELTAREPLACE | UPSERTSELECTTRUNCATE TABLE

Table of Content

6 2012-12-19

129129

130130132132133134134135135136137137138138139140140141141142142143143

144144144145146147

147147147148149

150150151153155155157157158159165

166167168169172

173173173174176178179180180

182184

185

UNLOADUPDATE

System Management StatementsALTER SYSTEM ALTER CONFIGURATIONALTER SYSTEM ALTER SESSION SETALTER SYSTEM ALTER SESSION UNSETALTER SYSTEM CANCEL [WORK IN] SESSIONALTER SYSTEM CLEAR SQL PLAN CACHEALTER SYSTEM CLEAR TRACESALTER SYSTEM DISCONNECT SESSIONALTER SYSTEM LOGGINGALTER SYSTEM PERSISTENCE ENCRYPTIONALTER SYSTEM RECLAIM DATAVOLUMEALTER SYSTEM RECLAIM LOGALTER SYSTEM RECLAIM VERSION SPACEALTER SYSTEM RECONFIGURE SERVICEALTER SYSTEM REMOVE TRACESALTER SYSTEM RESET MONITORING VIEWALTER SYSTEM SAVE PERFTRACEALTER SYSTEM SAVEPOINTALTER SYSTEM START PERFTRACEALTER SYSTEM STOP PERFTRACEALTER SYSTEM STOP SERVICESET SYSTEM LICENSEUNSET SYSTEM LICENSE ALL

Session Management StatementsCONNECTSET HISTORY SESSIONSET SCHEMASET [SESSION]UNSET [SESSION]

Transaction Management StatementsCOMMITLOCK TABLEROLLBACKSET TRANSACTION

Access Control StatementsALTER SAML PROVIDERALTER USERCREATE ROLECREATE SAML PROVIDERCREATE USERDROP ROLEDROP SAML PROVIDERDROP USERGRANTREVOKE

Data Import Export StatementsEXPORTIMPORTIMPORT FROMIMPORT SCAN

Procedural StatementsALTER PROCEDURE RECOMPILECALLCREATE FUNCTIONCREATE PROCEDURECREATE TYPEDROP FUNCTIONDROP PROCEDUREDROP TYPE

Restrictions for SQL StatementsSee Also

SQL Error Codes

Table of Content

7 2012-12-19

SAP HANA Database - SQL Reference Manual

This guide describes SQL language supported by SAP HANA DB.


8 2012-12-19

SQL Reference Manual


9 2012-12-19

Notation

This reference use BNF (Backus Naur Form) which is a notation technique used to define programminglanguages. BNF describes the syntax of a grammar using a set of production rules and a set of symbols.

Symbols used in BNF

Symbol Description

< >Angle brackets are used to surround the name of a syntactic element (BNF nontermina l ) of the SQLlanguage.

::=The defini tion operator i s used to provide defini tions of the element appeared on the left s ide ofthe operator in a production rule.

[ ]Square brackets are used to indicate optiona l elements in a formula . Optiona l elements may bespeci fied or omitted.

{ }Braces group elements in a formula . Repeti tive elements (zero or more elements ) can be speci fiedwithin brace symbols .

|The a l ternative operator indicates that the portion of the formula fol lowing the bar i s ana l ternative to the portion preceding the bar.

...The el l ips i s indicates that the element may be repeated any number of times . If el l ips i s appearsafter grouped elements speci fying that the grouped elements enclosed with braces are repeated.If el l ips i s appears a fter a s ingle element, only that element i s repeated.

!!Introduces normal Engl i sh text. This i s used when the defini tion of a syntactic element i s notexpressed in BNF.

BNF Lowest Terms Representations

Throughout the BNF used in this manual each syntax term will be defined to one of the lowest termrepresentations shown below.

<digit> ::= 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9

<letter> ::= a | b | c | d | e | f | g | h | i | j | k | l | m | n | o | p | q | r | s | t | u | v | w | x | y | z | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

<string_literal> ::= <double_quote><string_content><double_quote> | <single_quote><string_content><single _quote>

<string_content> = { <letter> | <digit> }...

<identifier> ::= <letter> { <letter> | <digit> }...

<password> ::= [{ <letter> | <digit> }...]

<sign> ::= + | -

<period> ::= .

<unsigned_integer> ::= <digit>...

<signed_integer> ::= [<sign>] <unsigned_integer>

<signed_numeric_literal> ::= [<sign>] <unsigned_numeric_literal>

<unsigned_numeric_literal> ::= <exact_numeric_literal> | <approximate_numeric_literal>

<exact_numeric_literal> ::= <unsigned_integer> [<period> [<unsigned_integer>]] | <period> <unsigned_integer>

<approximate_numeric_literal> ::= <mantissa> E <exponent>

<mantissa> ::= <exact_numeric_literal>

<exponent> ::= <signed_integer>


10 2012-12-19

Introduction

This chapter describes the SAP HANA Database implementation of Structured Query Language (SQL). Itexplains the characteristics of SQL, also how to manage comments and reserve words.

SQL

SQL stands for Structured Query Language. It is a standardized language for communicating with a relationaldatabase. SQL is used to retrieve, store or manipulate information in the database.

SQL statements perform the following tasks:

Schema definition and manipulationData manipulationSystem managementSession managementTransaction management

Supported Languages and Code Pages

The SAP HANA Database supports Unicode to allow the use of all languages in the Unicode Standard and 7 BitASCII code page without restriction.

Comment

You can add comments to improve readability and maintainability of your SQL statements. Comments aredelimited in SQL statements as follows:

Double hyphens "--". Everything after the double hyphen until the end of a line is considered by theSQL parser to be a comment"/*" and "*\/". This style of commenting is used to place comments on multiple lines. All text betweenthe opening "/*" and closing "*\/" is ignored by the SQL parser.

Identifiers

Syntax:

<identifier> ::= <simple_identifier> | <double_quotes><special_identifier><double_quotes> <simple_identifier> ::= <letter> [{<letter_or_digit>|<underscore>}, ...] <double_quotes> ::= " <special_identifier> ::= any character <letter> ::= A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | a | b | c | d | e | f | g | h | i | j | k | l | m | n | o | p | q | r | s | t | u | v | w | x | y | z <digit> ::= 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 <letter_or_digit> ::= <letter> | <digit> <underscore> ::= _

Identifiers are used to represent names used in SQL statement including table name, view name, synonymname, column name, index name, function name, procedure name, user name, role name, and so on. Thereare two kinds of identifiers; undelimited identifiers and delimited identifiers.

Undelimited table and column names must start with a letter and cannot contain any symbols otherthan digits or an underscore "_".Delimited identifiers are enclosed in the delimiter, double quotes, then the identifier can contain anycharacter including special characters. For example, "AB$%CD" is a valid identifier name.Limitations:

"_SYS_" is reserved exclusively for database engine, hence not allowed at the beginning ofschema object names.Role name and user name must be specified as undelimited identifiers.Maximum length for the identifiers is 127 characters.

Single Quotation Mark

Single quotation marks are used to delimit string literals and single quotation mark itself can be representedusing two single quotation marks.

Double Quotation Mark


11 2012-12-19

Double quotation marks are used to delimit identifiers and double quotation mark itself can be representedusing two double quotation marks.

SQL Reserved Words

Reserved words are words which have a special meaning to the SQL parser in the SAP HANA Database thatcannot be used as a user-defined name. Reserved words should not be used in SQL statements for schemaobject names. If necessary, you can work around this limitation by delimiting a table or column name withdouble quotation marks.

The following table lists all the current and future reserved words for the SAP HANA Database.

Table 1. Reserved Words

ALL ALTER AS BEFORE

BEGIN BOTH CASE CHAR

CONDITION CONNECT CROSS CUBE

CURRENT_CONNECTION CURRENT_DATE CURRENT_SCHEMA CURRENT_TIME

CURRENT_TIMESTAMP CURRENT_USER CURRENT_UTCDATE CURRENT_UTCTIME

CURRENT_UTCTIMESTAMP CURRVAL CURSOR DECLARE

DISTINCT ELSE ELSEIF ELSIF

END EXCEPT EXCEPTION EXEC

FOR FROM FULL GROUP

HAVING IF IN INNER

INOUT INTERSECT INTO IS

JOIN LEADING LEFT LIMIT

LOOP MINUS NATURAL NEXTVAL

NULL ON ORDER OUT

PRIOR RETURN RETURNS REVERSE

RIGHT ROLLUP ROWID SELECT

SET SQL START SYSDATE

SYSTIME SYSTIMESTAMP SYSUUID TOP

TRAILING UNION USING UTCDATE

UTCTIME UTCTIMESTAMP VALUES WHEN

WHERE WHILE WITH


12 2012-12-19

Data Types

This section describes the data types used in the SAP HANA Database.

Data type specifies the characteristics of a data value. A special value of NULL is included in every data typeto indicate the absence of a value. The following table shows the built-in data types available in the SAPHANA Database.

Classification of Data Types

In the SAP HANA Database each data type can be classified by its characteristic as follows:

Table 2: Classification of data types

Classification Data Type

Datetime types DATE, TIME, SECONDDATE, TIMESTAMP

Numeric types TINYINT, SMALLINT, INTEGER, BIGINT, SMALLDECIMAL, DECIMAL, REAL, DOUBLE

Character s tring types VARCHAR, NVARCHAR, ALPHANUM, SHORTTEXT

Binary types VARBINARY

Large Object types BLOB, CLOB, NCLOB, TEXT

Datetime Types

DATE The DATE data type consists of year, month, and day information to represent a date value. Thedefault format for the DATE data type is 'YYYY-MM-DD'. YYYY represents the year, MM representsthe month, and DD represents the day. The range of date value is 0001-01-01 through 9999-12-31.

TIME The TIME data type consists of hour, minute, and second to represent a time value. The defaultformat for the TIME data type is 'HH24:MI:SS'. HH24 represents the hour from 0 to 24, MI representsthe minute from 0 to 59, SS represents the second from 0 to 59.

SECONDDATE The SECONDDATE data type consists of year, month, day, hour, minute and second information torepresent a date with time value. The default format for the SECONDDATE data type is 'YYYY-MM-DDHH24:MI:SS'. YYYY represents the year, MM represents the month, DD represents the day, HH24represents hour, MI represents minute, and SS represents seconds. The range of date value is 0001-01-01 00:00:01 through 9999-12-31 24:00:00.

TIMESTAMP The TIMESTAMP data type consists of date and time information. Its default format is 'YYYY-MM-DDHH24:MI:SS.FF7'. FFn represents the fractional seconds where n indicates the number of digits infractional part. . The range of the timestamp value is 0001-01-01 00:00:00.0000000 through 9999-12-31 23:59:59.9999999.

For details on supported formats for datetime types, refer to Table 4, Table 5, Table 6 and Table 7 below.

Date Formats

The following date/time formats can be used when parsing a string into a date/time type and converting adate/time type value into a string value. Please note that format for Timestamp is the combination of Dateand Time with the additional support for fractional seconds.


13 2012-12-19

Table 4: Supported formats for Date

Format Description Examples

YYYY-MM-DD Defaul t format INSERT INTO my_tbl VALUES ('1957-06-13');

YYYY/MM/DD YYYY/MM-DD YYYY-MM/DD

YYYY from 0001 to 9999, MM from 1 to 12,DD from 1 to 31. If year has less than fourdigi ts , month has less than two digi ts , orday has less than two digi ts , then va lueswi l l be padded by one or more zeros . Forexample, a two digi t year 45 wi l l be savedas year 0045, a one digi t month 9 wi l l besaved as 09, and a one digi t day 2 wi l l besaved as 02.

INSERT INTO my_tbl VALUES ('1957-06-13'); INSERT INTO my_tbl VALUES ('1957/06/13'); INSERT INTO my_tbl VALUES ('1957/06-13'); INSERT INTO my_tbl VALUES ('1957-06/13');

YYYYMMDD ABAP Data Type, DATS format. INSERT INTO my_tbl VALUES ('19570613');

MON Abbreviated name of month. (JAN. ~ DEC.)

INSERT INTO my_tbl VALUES(TO_DATE('2040-Jan-10', 'YYYY-MON-DD')); INSERT INTO my_tbl VALUES(TO_DATE('Jan-10', 'MON-DD'));

MONTH Name of month. (JANUARY - DECEMBER).

INSERT INTO my_tbl VALUES(TO_DATE('2040-January-10', 'YYYY-MONTH-DD')); INSERT INTO my_tbl VALUES(TO_DATE('January-10', 'MONTH-DD'));

RM Roman numera l month (I-XII ; JAN = I).

INSERT INTO my_tbl VALUES(TO_DATE('2040-I-10', 'YYYY-RM-DD')); INSERT INTO my_tbl VALUES (TO_DATE('I-10', 'RM-DD'));

DDD Day of year (1-366).

INSERT INTO my_tbl VALUES(TO_DATE('204', 'DDD')); INSERT INTO my_tbl VALUES(TO_DATE('2001-204','YYYY-DDD'));

Time Formats

Table 5: Supported formats for Time


HH24:MI:SS Defaul t format

HH:MI[:SS][AM|PM] HH12:MI[:SS][AM|PM]HH24:MI[:SS]

HH from 0 to 23. MI from 0 to 59. SS from 0to 59. FFF from 0 to 999. If one digi t hour, minute, second i sspeci fied, then 0 wi l l be inserted intothe va lue. For example, 9:9:9 wi l l besaved as 09:09:09. HH12 indicates 12 hour clock and HH24indicates 24 hour clock.AM or PM can be speci fied as a suffix toindicate the time va lue i s before or a fternoon.

INSERT INTO my_tbl VALUES ('23:59:59'); INSERT INTO my_tbl VALUES ('3:47:39 AM'); INSERT INTO my_tbl VALUES ('9:9:9 AM'); INSERT INTO my_tbl VALUES(TO_TIME('11:59:59','HH12:MI:SS');

SSSSS Seconds past midnight (0-86399).INSERT INTO my_tbl VALUES(TO_TIME('12345', 'SSSSS'));

Timestamp Formats

Table 6: Supported formats for Timestamp


YYYY-MM-DDHH24:MI:SS.FF7

Defaul t format

FF [1..7]

Fractiona l seconds has the range 1 to 7after the FF parameter to speci fy thenumber of digi ts in the fractiona l secondportion of the date time va lue returned. Ifa digi t i s not speci fied, the defaul t va lueis used.

INSERT INTO my_tbl VALUES(TO_TIMESTAMP('2011-05-1112:59.999','YYYY-MM-DD HH:SS.FF3'));

Additional Formats


14 2012-12-19

Table 7: Additional formats for Datetime

Format Description Example

D Day of week (1-7). TO_CHAR(CURRENT_TIMESTAMP,'D')

DAY Name of day (MONDAY - SUNDAY). TO_CHAR(CURRENT_TIMESTAMP,'DAY')

DY Abbreviated name of day (MON - SUN). TO_CHAR(CURRENT_TIMESTAMP,'DY')

MON Abbreviated month name (JAN - DEC) TO_CHAR(CURRENT_TIMESTAMP,'MON')

MONTH Ful l month name (JANUARY - DECEMBER) TO_CHAR(CURRENT_TIMESTAMP,'MONTH')

RMRoman numera l month (I - XII ; I i s forJanuary)

TO_CHAR(CURRENT_TIMESTAMP,'RM')

Q Quarter of year (1, 2, 3, 4) TO_CHAR(CURRENT_TIMESTAMP,'Q')

W Week of month (1-5). TO_CHAR(CURRENT_TIMESTAMP,'W')

WW Week of year (1-53). TO_CHAR(CURRENT_TIMESTAMP,'WW')

Supported Functions for Date/Time types

ADD_DAYSADD_MONTHSADD_SECONDSADD_YEARSCOALESCECURRENT_DATECURRENT_TIMECURRENT_TIMESTAMPCURRENT_UTCDATECURRENT_UTCTIMECURRENT_UTCTIMESTAMPDAYNAMEDAYOFMONTHDAYOFYEARDAYS_BETWEENEXTRACTGREATESTGREATESTHOURIFNULLISOWEEKLAST_DAYLEASTLOCALTOUTCMINUTEMONTHMONTHNAMENEXT_DAYNULLIFQUARTERSECONDSECONDS_BETWEENTO_DATETO_DATSTO_TIMETO_TIMESTAMPUTCTOLOCALWEEKWEEKDAYYEAR

Numeric Types

TINYINT The TINYINT data type stores an 8-bit unsigned integer. The minimum value is 0 and the maximumvalue is 255 for TINYINT.

SMALLINT The SMALLINT data type stores a 16-bit signed integer. The minimum value is -32,768 and themaximum value is 32,767 for SMALLINT.

INTEGER The INTEGER data type stores a 32-bit signed integer. The minimum value is -2,147,483,648 and the


15 2012-12-19

maximum value is 2,147,483,647 for INTEGER.

BIGINT The BIGINT data type stores a 64-bit signed integer. The minimum value is -9,223,372,036,854,775,808 and the maximum value is 9,223,372,036,854,775,807 for BIGINT.

DECIMAL(precision, scale) or DEC(p,s) DECIMAL(p, s) is the SQL standard notation for fixed-point decimal. "p" specifies precision or thenumber of total digits (the sum of whole digits and fractional digits). "s" denotes scale or the numberof fractional digits. For example, if a column is defined as DECIMAL(5, 4), the numbers 3.14, 3.1415,3.141592 are stored in the column as 3.1400, 3.1415, 3.1416, respectively keeping the specifiedprecision(5) and scale(4).

The precision p, can range from 1 to 34. The scale can range from 0 to p. If the scale is not specified, itdefaults to 0.

When precision and scale are not specified, DECIMAL becomes a floating-point decimal number. Inthis case, precision and scale can vary within the range 1 to 34 for precision and -6,111 to 6,176 forscale depending on the stored value.

Examples: 0.0000001234 (1234 x 10-10) has the precision 4 and the scale 10. 1.0000001234(10000001234 x 10-10) has the precision 11 and scale 10. 1234000000 (1234x106) has the precision 4and scale -6.

When precision and scale are not specified, DECIMAL becomes a floating-point decimal number. Inthis case, precision and scale can vary within the range described above, 1~34 for precision and -6,111~6,176 for scale depending on the stored value.

SMALLDECIMAL The SMALLDECIMAL is a floating-point decimal number. The precision and scale can vary within therange, 1~16 for precision and -369~368 for scale depending on the stored value. SMALLDECIMAL issupported only on column store.

DECIMAL and SMALLDECIMAL are floating-point types. For instance, a decimal column can store anyof 3.14, 3.1415, 3.141592 whilst maintaining their precision.

DECIMAL(p, s) is the SQL standard notation for fixed-point decimal. For instance, 3.14, 3.1415,3.141592 are stored in a decimal(5, 4) column as 3.1400, 3.1415, 3.1416, respectively keeping thespecified precision(5) and scale(4).

REAL The REAL data type specifies a single-precision 32-bit floating-point number.

DOUBLE The DOUBLE data type specifies a single-precision 64-bit floating-point number. The minimum value is-1.79769 x 10308 and the maximum value is 1.79769x10308 . The smallest positive DOUBLE value is2.2207x10-308 and the largest negative DOUBLE value is -2.2207x10-308.

FLOAT(n) The FLOAT(n) data type specifies a 32-bit or 64-bit real number, where n specifies the number ofsignificant bits and can range between 1 and 53.

When you use the FLOAT(n) data type, if n is smaller than 25, it becomes a 32-bit REAL data type. If nis greater than or equal to 25, it then becomes a 64-bit DOUBLE data type. If n is not declared, itbecomes a 64-bit double data type by default.

Character String Types

The character string data types are used to store values that contain character strings. While VARCHAR datatypes contain ASCII character strings, NVARCHAR are used for storing Unicode character strings.

VARCHAR The VARCHAR(n) data type specifies a variable-length ASCII character string, where n indicates themaximum length and is an integer between 1 and 5000.

NVARCHAR The NVARCHAR(n) data type specifies a variable-length Unicode character set string, where nindicates the maximum length and is an integer between 1 and 5000.

ALPHANUM The ALPHANUM(n) data type specifies a variable-length character string which contains alpha-numeric characters, where n indicates the maximum length and is an integer between 1 and 127.


16 2012-12-19

SHORTTEXT The SHORTTEXT(n) data type specifies veriable-length character string which supports textsearch-features and stringsearch-features. This is not a standalone sql-type. Selecting a SHORTTEXT(n)-column yields a column of type NVARCHAR(n).

<shorttext_type> ::= SHORTTEXT '(' int_const ')' <elem_list_shorttext>

<elem_list_shorttext> ::= <elem_shorttext> [... ',' <elem_shorttext>]

<elem_shorttext> ::= <fulltext_elem> | SYNC[HRONOUS]

Binary Types

Binary types are used to store bytes of binary data.

VARBINARY The VARBINARY(n) data type is used to store binary data of a specified maximum length in bytes,where n indicates the maximum length and is an integer between 1 and 5000.

Large Object (LOB) Types

LOB (large objects) data types, CLOB, NCLOB and BLOB, are used to store a large amount of data such as textdocuments and images. The maximum size of an LOB is 2 GB.

BLOB The BLOB data type is used to store large binary data.

CLOB The CLOB data type is used to store large ASCII character data.

NCLOB The NCLOB data type is used to store a large Unicode character object.

TEXT The TEXT data type specifies which supports textsearch-features. This is not a standalone sql-type.Selecting a TEXT-column yields a column of type NCLOB.

<text_type> ::= TEXT <opt_fulltext_elem_list_text>

<opt_fulltext_elem_list_text> ::= <fulltext_elem_text> [... ',' <fulltext_elem_text>]

<fulltext_elem_text> ::= <fulltext_elem> | [SYNC[HRONOUS] | [ASYNC[HRONOUS] FLUSH [QUEUE] EVERY <n> MINUTES [[OR] AFTER <m> DOCUMENTS] ]

Syntax-rules common to TEXT and SHORTTEXT

<fulltext_elem> ::= LANGUAGE COLUMN <column_name> | LANGUAGE DETECTION '(' <str_const_list> ')' | MIME TYPE COLUMN <column_name> | FUZZY SEARCH INDEX [ON|OFF] | PHRASE INDEX RATIO [ON|OFF] | CONFIGURATION <str_const> | SEARCH ONLY [ON|OFF] | FAST PREPROCESS [ON|OFF]

LOB types are provided for storing and retrieving such large data. LOB types support the followingoperations.

The length() function returns the LOB length in bytes.LIKE can be used to search LOB columns.

The LOB types have the following restrictions:

LOB columns cannot appear in ORDER BY or GROUP BY clauses.LOB columns cannot appear in FROM clauses as a join predicate.LOB columns cannot appear in WHERE clauses as a predicate except LIKE, CONTAINS, =, or <>.LOB columns cannot appear in SELECT clauses as an aggregate function argument.LOB columns cannot appear in SELECT DISTINCT clauses.LOB columns cannot be used in set operations such as EXCEPT. UNION ALL is an exception.LOB columns cannot be used as a primary key.


17 2012-12-19

LOB columns cannot be used in CREATE INDEX statements.LOB columns cannot be used in statistics update statements.

Mapping between SQL Data Type and Column Store Data Type

SQL Type Column Store Type

Integer Types TINYINT, SMALLINT, INT CS_INT

BIGINT CS_FIXED(18,0)

Approximate Types REAL CS_FLOAT

DOUBLE CS_DOUBLE

FLOAT CS_DOUBLE

FLOAT(p) CS_FLOAT, CS_DOUBLE

Decimal Types DECIMAL CS_DECIMAL_FLOAT

DECIMAL(p,s ) CS_FIXED(p-s ,s )

SMALLDECIMAL CS_SDFLOAT

Character Types VARCHAR CS_STRING

NVARCHAR CS_STRING

CLOB, NCLOB CS_STRING

ALPHANUM CS_ALPHANUM

Binary Types BLOB CS_RAW

VARBINARY CS_RAW

Date/Time Types DATE CS_DAYDATE

TIME CS_SECONDTIME

TIMESTAMP CS_LONGDATE

SECONDDATE CS_SECONDDATE

Data Type Conversion

This section describes the data type conversion allowed in SAP HANA Database.

Explicit type conversion

The type of an expression result, for example a field reference, a function on fields, or literals can beconverted using the following functions: CAST, TO_ALPHANUM, TO_BIGINT, TO_VARBINARY,TO_BLOB, TO_CLOB, TO_DATE, TO_DATS, TO_DECIMAL, TO_DOUBLE, TO_INTEGER, TO_INT,TO_NCLOB, TO_NVARCHAR, TO_REAL, TO_SECONDDATE, TO_SMALLINT, TO_TINYINT, TO_TIME,TO_TIMESTAMP, TO_VARCHAR.

Implicit type conversion

When a given set of operand/argument types does not match what an operator/function expects, atype conversion is carried out by the SAP HANA Database. This conversion only occurs if a relevantconversion is available and if it makes the operation/function executable. For instance, a comparisonof BIGINT and VARCHAR is performed by implicitly converting VARCHAR to BIGINT. The entire explicitconversions can be used for implicit conversion except for the TIME and TIMESTAMP data types. TIMEand TIMESTAMP can be converted to each other using TO_TIME(TIMESTAMP) andTO_TIMESTAMP(TIME).

Examples

Table 8: Implicit Type conversion Examples

Input Expression Transformed Expression with Implicit Conversion

BIGINT > VARCHAR BIGINT > BIGINT(VARCHAR)

BIGINT > DECIMAL DECIMAL(BIGINT) > DECIMAL

TIMESTAMP > DATE TIMESTAMP > TIMESTAMP(DATE)

DATE > TIME Error because there i s no convers ion ava i lable between DATE and TIME

In the tables below,

Boxes with "OK" means data type conversions are allowed without any checks.Boxes with "CHK" means the data type can be converted if the data is valid for the target type.Boxes with "-" indicates that data type conversion is not allowed.


18 2012-12-19

The rules shown are applicable to both implicit and explicit conversion except for Time to Timestampconversion. Only explicit conversions are allowed for converting the Time data type to Timestamp using theTO_TIMESTAMP or CAST functions.

Table 9a: Data type conversion table

Target/Source

tinyint - OK OK OK OK OK OK OK OK OK OK

smallint CHK - OK OK OK OK OK OK OK OK OK

integer CHK CHK - OK OK OK OK OK OK OK OK

bigint CHK CHK CHK - OK CHK CHK CHK OK OK OK

decimal CHK CHK CHK CHK - CHK CHK CHK OK OK OK

decimal(p,s) CHK CHK CHK CHK CHK CHK CHK CHK CHK CHK OK

smalldecimal CHK CHK CHK CHK OK CHK - CHK CHK OK OK

real CHK CHK CHK CHK OK CHK CHK - OK OK OK

double CHK CHK CHK CHK CHK CHK CHK CHK - OK OK

varchar CHK CHK CHK CHK CHK CHK CHK CHK CHK - OK

nvarchar CHK CHK CHK CHK CHK CHK CHK CHK CHK CHK -

Table 9b: Data type conversion table

Target/Source

time date seconddate timestamp varchar nvarchar

time - - - - OK OK

date - - OK OK OK OK

seconddate time date - timestamp OK OK

timestamp time date seconddate - OK OK

varchar CHK CHK CHK CHK - OK

nvarchar CHK CHK CHK CHK CHK -

Table 9c: Data type conversion table

Target/Source

varbinary alphanum varchar nvarchar

varbinary - - - -

alphanum - - OK OK

varchar OK OK - OK

nvarchar OK OK CHK -

Data Type Precedence

This section describes the data type precedence implemented by the SAP HANA Database. Data typeprecedence specifies that the data type with lower precedence is converted to the data type with higherprecedence.

tinyi

nt

smal

lint

inte

ger

bigi

nt

deci

mal

deci

mal

(p,s

)

smal

ldec

imal

real

doub

le

varc

har

nvar

char


19 2012-12-19

Highest TIMESTAMP

SECONDDATE

DATE

TIME

DOUBLE

REAL

DECIMAL

SMALLDECIMAL

BIGINT

INTEGER

SMALLINT

TINYINT

NCLOB

NVARCHAR

CLOB

VARCHAR

BLOB

Lowest VARBINARY

Typed Constant

A constant is a symbol that represents a specific fixed data value.

Character string contant A character string constant is enclosed in single quotation marks.

'Brian''100'

Unicode string has a similar format to character string but is preceded by an N identifier (N stands forNational Language in the SQL-92 standard). The N prefix must be uppercase.

N'abc'

SELECT 'Brian' "character string 1", '100' "character string 2", N'abc' "unicode string" FROM DUMMY;

character string 1, character string 2, unicode string Brian, 100, abc

Number constant A number constant is represented by a string of numbers that are not enclosed in quotation marks.Numbers may contain a decimal point or scientific notation.

123123.41.234e2

A hexadecimal number constant is a string of hexadecimal numbers and has the prefix 0x.

0x0abc

SELECT 123 "integer", 123.4 "decimal1", 1.234e2 "decimal2", 0x0abc "hexadecimal" FROM DUMMY;

integer, decimal1, decimal2, hexadecimal 123, 123.4, 123.4, 2748

Binary string constant A binary string has the prefix X and is a string of hexadecimal numbers that are enclosed in quotationmarks.

X'00abcd'x'dcba00'

SELECT X'00abcd' "binary string 1", x'dcba00' "binary string 2" FROM DUMMY;

binary string 1, binary string 2


20 2012-12-19

00ABCD, DCBA00

Date/Time/Timestamp constant Date, Time and Timestamp each have the following prefixes.

date'2010-01-01'time'11:00:00.001'timestamp'2011-12-31 23:59:59'

SELECT date'2010-01-01' "date", time'11:00:00.001' "time", timestamp'2011-12-31 23:59:59' "timestamp" FROM DUMMY;

date, time, timestamp 2010-01-01, 11:00:00, 2011-12-31 23:59:59.0


21 2012-12-19

Predicates

A predicate is specified by combining one or more expressions, or logical operators, and returns one of thefollowing logical/truth values: TRUE, FALSE, or UNKNOWN.

Comparison Predicates

Two values are compared using comparison predicates and the comparison returns true, false, or unknown.

Syntax:

<comparison_predicate> ::= <expression> { = | != | <> | > | < | >= | <= } [ ANY | SOME| ALL ] { <expression_list> | <subquery> } <expression_list> ::= <expression>, ...

Expressions can be a simple expression such as a character, date, or number. An expression can also be ascalar subquery.

ANY, SOME - When ANY or SOME are specified, the comparison returns true if at least one value returned bythe subquery or expression_list is true. ALL - When ALL is specified, the comparison returns true if thecomparison of all values returned by the subquery or expression_list is true.

Range Predicate

A value is compared with a list of values within the provided range.

Syntax:

<range_predicate> ::= <expression1> [NOT] BETWEEN <expression2> AND <expression3>

BETWEEN ... AND ... - When a range predicate is used, it returns true if expression1 is within the rangespecified by expression2 and expression3. True will only be returned if expression2 has a lesser value thanexpression3.

In Predicate

A value is compared with a specified set of values. True will be returned if the value of expression1 is found inthe expression_list (or subquery).

Syntax:

<in_predicate> ::= <expression> [NOT] IN { <expression_list> | <subquery> }

Exists Predicate

Returns true if the subquery returns a result set that is not empty and returns false if the subquery returns anempty result set.

Syntax:

<exists_predicate> ::= [NOT] EXISTS ( <subquery> )

LIKE Predicate

The LIKE predicate is used for string comparisons. Expression1 is tested for a pattern contained inexpression2. Wildcard characters ( % ) and ( _ ) may be used in the comparison string expression2. LIKEreturns true if the pattern specified by expression2 is found.

The percentage sign (%) matches zero or more characters and underscore (_) matches exactly onecharacter. To match a percent sign or underscore in the LIKE predicate, an escape character must be used.

Using the optional argument, ESCAPE expression3, you can specify the escape character that will be usedallowing the underscore (_) or percentage sign (%) to be matched.

Syntax:

<like_predicate> ::= <expression1> [NOT] LIKE <expression2> [ESCAPE <expression3>]

NULL Predicate


22 2012-12-19

When the IS NULL predicate is specified, a value can be compared with NULL. IS NULL returns true if theexpression value is NULL. If the IS NOT NULL predicate is specified, it returns true if a value is not NULL.

Syntax:

<null_predicate> ::= <expression> IS [NOT] NULL

CONTAINS Predicate

The CONTAINS predicate is used to search for text-matches to a search string in subqueries.

Syntax:

<contains_function> ::= CONTAINS '(' <contains_columns> ',' <search_string>')' | CONTAINS '(' <contains_columns> ',' <search_string> ',' <search_specifier> ')' <contains_columns> ::= '*' | <column_name> | '(' <columnlist> ')' <search_string> ::= <string_const> <search_specifier> ::= <search_type> <opt_search_specifier2_list> | <search_specifier2_list> <opt_search_specifier2_list> ::= empty | <search_specifier2_list> <search_type> ::= <exact_search> | <fuzzy_search> | <linguistic_search> <search_specifier2_list> ::= <search_specifier2> | <search_specifier2_list> ',' <search_specifier2> <search_specifier2> := <weights> | <language> <exact_search> ::= EXACT <fuzzy_search> ::= FUZZY | FUZZY '(' <float_const> ')' | FUZZY '(' <float_const> ',' <additional_params> ')' <linguistic_search> ::= LINGUISTIC <weights> ::= WEIGHT '(' <float_const_list> ')' <language> :: LANGUAGE '(' <string_const> ')' <additional_params> ::= <string_const>

search_stringThe freestyle-search-string format is used (eg. Peter "Palo Alto" OR Berlin -"SAP LABS" )

search_specifierIf the search_specifier clause is not specified EXACT is taken as default

EXACTEXACT returns true for those records where exact matches of the searchterms are found in the search-attributes

FUZZYFUZZY returns true for those records where words similar to the searchterms are found in the search-attributes (e.g. spelling errors will be ignored to a certain extent)

float_constIf float_const is omitted, then 0.8 is the default. This default can be overridden by defining parameterFUZZINESSTHRESHOLD supported by columnstore join-views

WEIGHTIf a weights list is specified, it must be the same size as the number of (expanded) columns in<contains_columns>

LANGUAGE


23 2012-12-19

LANGUAGE is used during preprocessing of the search string and as a pre-search filter. Only documents whichmatch the search string and the language specified are returned.

LINGUISTICLINGUISTIC returns true for those records where word-variants of the searchterms are found in the search-attributes (e.g. searching for 'cats' will also return records which contain 'cat')

Limitations: If there are multiple CONTAINS predicates specified in the where clause of a select statement,then only one of the predicates may consist of more than one column in the <contains_columns> list

The CONTAINS-predicate works only on column store tables (simple tables and join-views)

Examples:

Exact search

select * from T where contains(column1, 'dog OR cat') -- EXACT is implicit select * from T where contains(column1, 'dog OR cat', EXACT) select * from T where contains(column1, '"cats and dogs"') -- phrase search

Fuzzy search

select * from T where contains(column1, 'catz', FUZZY(0.8))

Linguistic search:

select * from T where contains(column1, 'catz', LINGUISTIC)

Freestyle search: The freestyle-search is a search mulitple columns.

select * from T where CONTAINS( (column1,column2,column3), 'cats OR dogz', FUZZY(0.7)) select * from T where CONTAINS( (column1,column2,column3), 'cats OR dogz', FUZZY(0.7))


24 2012-12-19

Operators

You can perform arithmetic operations in expressions by using operators. Operators can be used forcalculation, value comparison or to assign values.

Unary and Binary Operators

Table 10. Unary and binary operators

Operator Operation Format Description

UnaryA unary operator appl ies to one operand or as ingle va lue express ion.

operator operand

unary plusoperator(+)

unary negationoperator(-)

logica lnegation(NOT)

BinaryBinary A binary operator appl ies to two operandsor two va lue express ions .

operand1 operatoroperand2

multipl i cativeoperators ( *, / )

addi tive operators (+,- )

comparisonoperators ( =,!=,<,>,<=,>=)

logica l operators (AND, OR )

Operator Precedence

An expression can use several operators. If the number of operators is greater than one then the SAP HANADatabase will evaluate them in order of operator precedence. You can change the order of evaluation byusing parentheses, as expressions contained within parentheses are always evaluated first.

If parentheses are not used, the operators have the precedence indicated by the table below. Please note,the SAP HANA Database will evaluate operators with equal precedence from left to right within anexpression.

Table 11. SQL operator precedence

Precedence Operator Operation

Highest () parentheses

+, - unary pos i tive and negative operation

*, / multipl i cation, divi s ion

+, - addi tion, subtraction

|| concatenation

=, !=, <, >, <=, >=, IS NULL, LIKE, BETWEEN compars ion

NOT logica l negation

AND conjunction

Lowes t OR dis junction

Arithmetic Operators

You use arithmetic operators to perform mathematical operations such as adding, subtracting, multiplying,dividing and negation of numeric values.


25 2012-12-19

Table 12. Arithmetic operators

Operator Description

-<express ion> Negation. If the express ion i s the NULL va lue, the resul t i s NULL.

<express ion> +<express ion>

Addition. If ei ther express ion i s the NULL va lue, the resul t i s NULL.

<express ion> -<express ion>

Subtraction. If ei ther express ion i s the NULL va lue, the res ul t i s NULL.

<express ion> *<express ion>

Multipl i cation. If ei ther express ion i s NULL, the resul t i s NULL.

<express ion> /<express ion>

Divis ion. If ei ther express ion i s NULL, or i f the s econd express ion i s 0, an error i sreturned.

String Operators

A concatenation operator combines two items such as strings, expressions or constants into one.

Table 13. Concatenation operators

Operator Description

<express ion> || <express ion>String concatenation (two vertica l bars ).If ei ther s tring i s NULL, i t returns NULL.

For VARCHAR or NVARCHAR type strings, leading or trailing spaces are kept. If either string is of data typeNVARCHAR, the result has data type NVARCHAR and is limited to 5000 characters. The maximum length forVARCHAR concatenation is also limited to 5000 characters.

Comparsion Operators

Syntax:

<comparison_operation> ::= <expression1> <comparison_operator> <expression2>

Table 14. Comparison operators

Operator Description Example

= Equa l to SELECT * FROM students WHERE id = 25;

> Greater than SELECT * FROM students WHERE id > 25;

< Less than SELECT * FROM students WHERE id < 25;

>= Greater than or equa l to SELECT * FROM students WHERE id >= 25;

<= Less than or equa l to SELECT * FROM students WHERE id <= 25;

!=, <> Not equa lSELECT * FROM students WHERE id != 25; SELECT * FROM students WHERE id <> 25;

Logical Operators

Search conditions can be combined using AND or OR operators. You can also negate them using the NOToperator.

Table 15. Logical operators

Operator Syntax Description

AND

WHEREcondition1ANDcondition2

When us ing AND, the combined condi tion i s TRUE i f both condi tions are TRUE,FALSE i f ei ther condi tion i s FALSE, and UNKNOWN otherwise.

OR

WHEREcondition1ORcondition2

When us ing OR, the combined condi tion i s TRUE i f ei ther condi tion i s TRUE,FALSE i f both condi tions are FALSE, and UNKNOWN otherwise.

NOTWHERE NOTcondition

The NOT operator i s placed before a condi tion to negate the condi tion. The NOTcondition i s TRUE i f condi tion i s FALSE, FALSE i f condi tion i s TRUE, and UNKNOWNif condi tion i s UNKNOWN.

Set Operators

The operators described in this section perform set operations on the results of two or more queries.


26 2012-12-19

Table 16. Set operators

Operator Returned Value

UNION Combines the resul ts of two or more select s tatements or query expres s ions

UNIONALL

Combines the resul ts of two or more select s tatements or query expres s ions , including a l ldupl icate rows .

INTERSECTCombines the resul ts of two or more select s tatements or query expres s ions , and returns a l lcommon rows .

EXCEPT Takes output from the fi rs t query and then removes rows selected by the second query.


27 2012-12-19

Expressions

An expression is a clause that can be evaluated to return values.

Syntax:

<expression> ::= <case_expression> | <function_expression> | <aggregate_expression> | (<expression> ) | ( <subquery> ) | - <expression> | <expression> <operator> <expression> | <variable_name> | <constant> | [<correlation_name>.]<column_name>

Case Expressions

A case expression allows the user to use IF ... THEN ... ELSE logic without using procedures in SQL statements.

Syntax:

<case_expression> ::= CASE <expression> WHEN <expression> THEN <expression> ... [ ELSE <expression>] { END }

If the expression following the CASE statement is equal to the expression following the WHEN statement,then the expression following the THEN statement is returned. Otherwise the expression following the ELSEstatement is returned if it exists.

Function Expressions

SQL built-in functions can be used as an expression.

Syntax:

<function_expression> ::= <function_name> ( <expression>, ... )

Aggregate Expressions

An aggregate expression uses an aggregate function to calculate a single value from the values of multiplerows in a column.

Syntax:

<aggregate_expression> ::= COUNT(*) | <agg_name> ( [ ALL | DISTINCT ] <expression> ) <agg_name> ::= COUNT | MIN | MAX | SUM | AVG | STDDEV | VAR

Aggregatename

Description

COUNTCounts the number of rows returned by a query. COUNT(*) returns the number of rows , regardlessof the va lue of those rows and including dupl icate va lues . COUNT(<express ion>) returns thenumber of non-NULL va lues for that express ion returned by the query.

MIN Returns the minimum va lue of express ion.

MAX Returns the maximum va lue of express ion.

SUM Returns the sum of express ion.

AVG Returns the ari thmetica l mean of express ion.

STDDEV Returns the s tandard deviation of given express ion as the square root of VARIANCE function.

VAR Returns the variance of express ion as the square of s tandard deviation.

Subqueries in Expressions


28 2012-12-19

A subquery is a SELECT statement enclosed in parentheses. The SELECT statement can contain one and onlyone select list item. When used as an expression a scalar subquery is allowed to return only zero or one value.

Syntax:

<scalar_subquery_expression> ::= (<subquery>)

Within the SELECT list of the top level SELECT, or in the SET clause of an UPDATE statement, you can use ascalar subquery anywhere that you can use a column name. However, scalar_subquery cannot be used insideGROUP BY clause.

Example:

For example, the following statement returns the number of employees in each department, grouped bydepartment name:

SELECT DepartmentName, COUNT(*), 'out of', (SELECT COUNT(*) FROM Employees) FROM Departments AS D, Employees AS E WHERE D.DepartmentID = E.DepartmentID GROUP BY DepartmentName;


29 2012-12-19

SQL Functions

Introduction

This chapter describes SQL Functions that are provided by SAP HANA Database.

Data Type Conversion FunctionsDateTime FunctionsNumber FunctionsString FunctionsWindow FunctionsMiscellaneous Functions

Data Type Conversion Functions

Data type conversion functions are used to convert arguments from one data type to another, or to testwhether a conversion is possible.

CAST

Syntax

CAST (expression AS data_type)

Syntax Elements

expression - The expression to be converted. data type - The target data type. TINYINT | SMALLINT |INTEGER | BIGINT | DECIMAL | SMALLDECIMAL | REAL | DOUBLE | ALPHANUM | VARCHAR | NVARCHAR |DAYDATE | DATE | TIME | SECONDDATE | TIMESTAMP

Description

Returns the value of an expression converted to a supplied data type.

Example

SELECT CAST (7 AS VARCHAR) "cast" FROM DUMMY; cast 7

TO_ALPHANUM

Syntax

TO_ALPHANUM (value)

Description

Converts a given value to an ALPHANUM data type.

Example

SELECT TO_ALPHANUM ('10') "to alphanum" FROM DUMMY; to alphanum 10

TO_BIGINT

Syntax

TO_BIGINT (value)

Description

Converts a value to a BIGINT data type.


30 2012-12-19

Example

SELECT TO_BIGINT ('10') "to bigint" FROM DUMMY; to bigint 10

TO_BINARY

Syntax

TO_BINARY (value)

Description

Converts a value to a BINARY data type.

Example

SELECT TO_BINARY ('abc') "to binary" FROM DUMMY; to binary 616263

TO_BLOB

Syntax

TO_BLOB (value)

Description

Converts a value to a BLOB data type. The argument value must be a binary string.

Example

SELECT TO_BLOB (TO_BINARY('abcde')) "to blob" FROM DUMMY; to blob abcde

TO_CLOB

Syntax

TO_CLOB (value)

Description

Converts a value to a CLOB data type.

Example

SELECT TO_CLOB ('TO_CLOB converts the value to a CLOB data type') "to clob" FROM DUMMY; to clob TO_CLOB converts the value to a CLOB data type

TO_DATE

Syntax

TO_DATE (d [, format])

Description


31 2012-12-19

Converts a date string d into a DATE data type. If the format specifier is omitted, the conversion is performedusing the date format model as explained in Date Formats.

Example

SELECT TO_DATE('2010-01-12', 'YYYY-MM-DD') "to date" FROM DUMMY; to date 2010-01-12

TO_DATS

Syntax

TO_DATS (d)

Description

Converts a date string d into an ABAP DATE string with format 'YYYYMMDD'.

Example

SELECT TO_DATS ('2010-01-12') "abap date" FROM DUMMY; abap date 20100112

TO_DECIMAL

Syntax

TO_DECIMAL (value [, precision, scale])

Description

Converts the value to a DECIMAL data type.

The precision is the total number of significant digits and can range from 1 to 34.

The scale is the number of digits from the decimal point to the least significant digit and can range from -

6,111 to 6,176. This means that the scale specifies the range of the exponent in the decimal number from 10-

6111 to 106176. If the scale is not specified, it defaults to 0. Scale is positive when the number has significantdigits to the right of the decimal point, and negative when the number has significant digits to the left of thedecimal point.

When precision and scale are not specified, DECIMAL becomes a floating-point decimal number. In this case,precision and scale can vary within the range described above, 1~34 for precision and -6,111~6,176 for scaledepending on the stored value.

Example

SELECT TO_DECIMAL(7654321.888888, 10, 3) "to decimal" FROM DUMMY

to decimal 7654321.889

TO_DOUBLE

Syntax

TO_DOUBLE (value)

Description

Converts a value to a DOUBLE (double precision) data type.


32 2012-12-19

Example

SELECT 3*TO_DOUBLE ('15.12') "to double" FROM DUMMY; to double 45.36

TO_INT

Syntax

TO_INT (value)

Description

Converts the value to an INTEGER data type.

Example

SELECT TO_INT('10') "to int" FROM DUMMY; to int 10

TO_INTEGER

Syntax

TO_INTEGER (value)

Description

Converts the value to an INTEGER data type.

Example

SELECT TO_INTEGER ('10') "to int" FROM DUMMY; to int 10

TO_NCLOB

Syntax

TO_NCLOB (value)

Description

Converts a value to a NCLOB data type.

Example

SELECT TO_NCLOB ('TO_NCLOB converts the value to a NCLOB data type') "to nclob" FROM DUMMY;

to nclob TO_NCLOB converts the value to a NCLOB data type

TO_NVARCHAR

Syntax

TO_NVARCHAR (value [,format])

Description


33 2012-12-19

Converts the value to a NVARCHAR unicode character data type. If the format specifier is omitted, theconversion is performed using the date format model as explained in Date Formats.

Example

SELECT TO_NVARCHAR(TO_DATE('2009/12/31'), 'YY-MM-DD') "to nchar" FROM DUMMY; to nchar 09-12-31

TO_REAL

Syntax

TO_REAL (value)

Description

Converts a value to a REAL (single precision) data type.

Example

SELECT 3*TO_REAL ('15.12') "to real" FROM DUMMY; to real 45.36000061035156

TO_SECONDDATE

Syntax

TO_SECONDDATE (d [, format])

Description

Converts a date string d into a SECONDDATE data type. If the format specifier is omitted, the conversion isperformed using the date format model as explained in Date Formats.

Example

SELECT TO_SECONDDATE ('2010-01-11 13:30:00', 'YYYY-MM-DD HH24:MI:SS') "to seconddate" FROM DUMMY; to seconddate 2010-01-11 13:30:00.0

TO_SMALLDECIMAL

Syntax

TO_SMALLDECIMAL (value)

Description

Converts the value to a SMALLDECIMAL data type.

Example

SELECT TO_SMALLDECIMAL(7654321.89) "to smalldecimal" FROM DUMMY;

to smalldecimal 7654321.89

TO_SMALLINT

Syntax


34 2012-12-19

TO_SMALLINT (value)

Description

Converts the value to a SMALLINT data type.

Example

SELECT TO_SMALLINT ('10') "to smallint" FROM DUMMY; to smallint 10

TO_TIME

Syntax

TO_TIME (t [, format])

Description

Converts a time string t into the TIME data type. If the format specifier is omitted, the conversion isperformed using the time format model as explained in Time Formats.

Example

SELECT TO_TIME ('08:30 AM', 'HH:MI AM') "to time" FROM DUMMY; to time 08:30:00

TO_TIMESTAMP

Syntax

TO_TIMESTAMP (d [, format])

Description

Converts a date string d into the TIMESTAMP data type. If the format specifier is omitted, the conversion isperformed using the date format model as explained in Date Formats.

Example

SELECT TO_TIMESTAMP ('2010-01-11 13:30:00', 'YYYY-MM-DD HH24:MI:SS') "to timestamp" FROM DUMMY; to timestamp 2010-01-11 13:30:00.0

TO_TINYINT

Syntax

TO_TINYINT (value)

Description

Converts the value to a TINYINT data type.

Example

SELECT TO_TINYINT ('10') "to tinyint" FROM DUMMY; to tinyint 10


35 2012-12-19

TO_VARCHAR

Syntax

TO_VARCHAR (value [, format])

Description

Converts a given value to a VARCHAR character data type. If the format specifier is omitted, the conversion isperformed using the date format model as explained in Date Formats.

Example

SELECT TO_VARCHAR (TO_DATE('2009-12-31'), 'YYYY/MM/DD') "to char" FROM DUMMY; to char 2009/12/31

DateTime Functions

ADD_DAYS

Syntax

ADD_DAYS (d, n)

Description

Computes the date d plus n days.

Example

SELECT ADD_DAYS (TO_DATE ('2009-12-05', 'YYYY-MM-DD'), 30) "add days" FROM DUMMY; add days 2010-01-04

ADD_MONTHS

Syntax

ADD_MONTHS (d, n)

Description

Computes the date d plus n months.

Example

SELECT ADD_MONTHS (TO_DATE ('2009-12-05', 'YYYY-MM-DD'), 1) "add months" FROM DUMMY; add months 2010-01-05

ADD_SECONDS

Syntax

ADD_SECONDS (t, n)

Description

Computes the time t plus n seconds.

Example


36 2012-12-19

SELECT ADD_SECONDS (TO_TIMESTAMP ('2012-01-01 23:30:45'), 60*30) "add seconds" FROM DUMMY; add seconds 2012-01-02 00:00:45.0

ADD_YEARS

Syntax

ADD_YEARS (d, n)

Description

Computes the date d plus n years.

Example

SELECT ADD_YEARS (TO_DATE ('2009-12-05', 'YYYY-MM-DD'), 1) "add years" FROM DUMMY; add years 2010-12-05

CURRENT_DATE

Syntax

CURRENT_DATE

Description

Returns the current local system date.

Example

SELECT CURRENT_DATE "current date" FROM DUMMY; current date 2010-01-11

CURRENT_TIME

Syntax

CURRENT_TIME

Description

Returns the current local system time.

Example

SELECT CURRENT_TIME "current time" FROM DUMMY; current time 17:37:37

CURRENT_TIMESTAMP

Syntax

CURRENT_TIMESTAMP

Description

Returns the current local system timestamp information.


37 2012-12-19

Example

SELECT CURRENT_TIMESTAMP "current timestamp" FROM DUMMY; current timestamp 2010-01-11 17:38:48.802

CURRENT_UTCDATE

Syntax

CURRENT_UTCDATE

Description

Returns the current UTC date. The UTC stands for Coordinated Universal Time, also known as GreenwichMean Time (GMT).

Example

SELECT CURRENT_UTCDATE "Coordinated Universal Date" FROM DUMMY; Coordinated Universal Time 2010-01-11

CURRENT_UTCTIME

Syntax

CURRENT_UTCTIME

Description

Returns the current UTC time.

Example

SELECT CURRENT_UTCTIME "Coordinated Universal Time" FROM DUMMY; Coordinated Universal Time 08:41:19

CURRENT_UTCTIMESTAMP

Syntax

CURRENT_UTCTIMESTAMP

Description

Returns the current UTC timestamp.

Example

SELECT CURRENT_UTCTIMESTAMP "Coordinated Universal Timestamp" FROM DUMMY; Coordinated Universal Timestamp 2010-01-11 08:41:42.484

DAYNAME

Syntax

DAYNAME (d)

Description


38 2012-12-19

Returns the weekday in English for date d.

Example

SELECT DAYNAME ('2011-05-30') "dayname" FROM DUMMY; dayname MONDAY

DAYOFMONTH

Syntax

DAYOFMONTH (d)

Description

Returns an integer the day of the month for date d.

Example

SELECT DAYOFMONTH ('2011-05-30') "dayofmonth" FROM DUMMY; dayofmonth 30

DAYOFYEAR

Syntax

DAYOFYEAR (d)

Description

Returns an integer representation of the day of the year for date d.

Example

SELECT DAYOFYEAR ('2011-05-30') "dayofyear" FROM DUMMY; dayofyear 150

DAYS_BETWEEN

Syntax

DAYS_BETWEEN (d1, d2)

Description

Computes the number of days between d1 and d2.

Example

SELECT DAYS_BETWEEN (TO_DATE ('2009-12-05', 'YYYY-MM-DD'), TO_DATE('2010-01-05', 'YYYY-MM-DD')) "days between" FROM DUMMY; days between 31

EXTRACT

Syntax

EXTRACT ({YEAR | MONTH | DAY | HOUR | MINUTE | SECOND} FROM d)


39 2012-12-19

Description

Finds and returns the value of a specified datetime field from date d.

Example

SELECT EXTRACT (YEAR FROM TO_DATE ('2010-01-04', 'YYYY-MM-DD')) "extract" FROM DUMMY; extract 2010

HOUR

Syntax

HOUR (t)

Description

Returns an integer representation of the hour for time t.

Example

SELECT HOUR ('12:34:56') "hour" FROM DUMMY; hour 12

ISOWEEK

Syntax

ISOWEEK (d)

Description

Returns the ISO year and week numbers of date d. The week number is prefixed by the letter W.

Please also see WEEK

Example

SELECT ISOWEEK (TO_DATE('2011-05-30', 'YYYY-MM-DD')) "isoweek" FROM DUMMY; isoweek 2011-W22

LAST_DAY

Syntax

LAST_DAY (d)

Description

Returns the date of the last day of the month that contains the date d.

Example

SELECT LAST_DAY (TO_DATE('2010-01-04', 'YYYY-MM-DD')) "last day" FROM DUMMY; last day 2010-01-31

LOCALTOUTC


40 2012-12-19

Syntax

LOCALTOUTC (t, timezone)

Description

Convert the local time t from a timezone to the UTC(GMT) time

Example

SELECT LOCALTOUTC (TO_TIMESTAMP('2012-01-01 01:00:00', 'YYYY-MM-DD HH24:MI:SS'), 'EST') "localtoutc" FROM DUMMY; localtoutc 2012-01-01 06:00:00.0

MINUTE

Syntax

MINUTE (t)

Description

Returns an integer representation of the minute for time t.

Example

SELECT MINUTE ('12:34:56') "minute" FROM DUMMY; minute 34

MONTH

Syntax

MONTH(d)

Description

Returns the number of the month from date d.

Example

SELECT MONTH ('2011-05-30') "month" FROM DUMMY; month 5

MONTHNAME

Syntax

MONTHNAME(d)

Description

Returns the name of the month in English for date d.

Example

SELECT MONTHNAME ('2011-05-30') "monthname" FROM DUMMY; monthname MAY


41 2012-12-19

NEXT_DAY

Syntax

NEXT_DAY (d)

Description

Returns the date of the next day after date d.

Example

SELECT NEXT_DAY (TO_DATE ('2009-12-31', 'YYYY-MM-DD')) "next day" FROM DUMMY; next day 2010-01-01

NOW

Syntax

NOW ()

Description

Returns the current timestamp.

Example

SELECT NOW () "now" FROM DUMMY; now 2010-01-01 16:34:19.894

QUARTER

Syntax

QUARTER (d, [, start_month ])

Description

Returns the numerical year quarter of date d. The first quarter starts in the month specified by start_month.If start_month is not specified the first quarter is assumed to begin in January.

Example

SELECT QUARTER (TO_DATE('2012-01-01', 'YYYY-MM-DD'), 2) "quarter" FROM DUMMY; quarter 2011-Q4

SECOND

Syntax

SECOND (t)

Description

Returns an integer representation of the second for time t.

Example

SELECT SECOND ('12:34:56') "second" FROM DUMMY; second


42 2012-12-19

second 56

SECONDS_BETWEEN

Syntax

SECONDS_BETWEEN (d1, d2)

Description

Computes the number of seconds between date arguments d1 and d2, which is semantically equal to d2 - d1.

Example

SELECT SECONDS_BETWEEN ('2009-12-05', '2010-01-05') "seconds between" FROM DUMMY; Seconds between 2678400

UTCTOLOCAL

Syntax

UTCTOLOCAL (t, timezone)

Description

Convert the UTC(GMT) time t to the local time in a timezone

Example

SELECT UTCTOLOCAL (TO_TIMESTAMP('2012-01-01 01:00:00', 'YYYY-MM-DD HH24:MI:SS'), 'EST') "utctolocal" FROM DUMMY; utctolocal 2011-12-31 20:00:00.0

WEEK

Syntax

WEEK (d)

Description

Returns the week number of date d.

Please also see ISOWEEK

Example

SELECT WEEK (TO_DATE('2011-05-30', 'YYYY-MM-DD')) "week" FROM DUMMY; week 23

WEEKDAY

Syntax

WEEKDAY (d)

Description

Returns an integer representation of the day of the week for date d. The return value ranges from 0 to 6,representing Monday(0) through to Sunday(6).


43 2012-12-19

Example

SELECT WEEKDAY (TO_DATE ('2010-12-31', 'YYYY-MM-DD')) "week day" FROM DUMMY; week day 4

YEAR

Syntax

YEAR (d)

Description

Returns the year number of date d.

Example

SELECT YEAR (TO_DATE ('2011-05-30', 'YYYY-MM-DD')) "year" FROM DUMMY; year 2011

Number Functions

Number functions take numeric values, or strings with numeric characters, as inputs and return numericvalues. When strings with numeric characters are provided as inputs implicit conversion from a string to anumber is performed automatically before results are computed.

ABS

Syntax

ABS (n)

Description

Returns the absolute value of the numeric argument n.

Example

SELECT ABS (-1) "absolute" FROM DUMMY; absolute 1

ACOS

Syntax

ACOS (n)

Description

Returns the arc-cosine, in radians, of the numeric argument n between -1 and 1.

Example

SELECT ACOS (0.5) "acos" FROM DUMMY; acos 1.0471975511965979

ASIN

Syntax


44 2012-12-19

ASIN (n)

Description

Returns the arc-sine, in radians, of the numeric argument n between -1 and 1.

Example

SELECT ASIN (0.5) "asin" FROM DUMMY; asin 0.5235987755982989

ATAN

Syntax

ATAN (n)

Description

Returns the arc-tangent, in radians, of the numeric argument n. The range of n is unlimited.

Example

SELECT ATAN (0.5) "atan" FROM DUMMY; atan 0.4636476090008061

ATAN2

Syntax

ATAN2 (n, m)

Description

Returns the arc-tangent, in radians, of the ratio of two numbers n and m. This produces the same result asATAN(n/m).

Example

SELECT ATAN2 (1.0, 2.0) "atan2" FROM DUMMY; atan2 0.4636476090008061

BINTOHEX

Syntax

BINTOHEX (expression)

Description

Converts a binary value to a hexadecimal value.

Example

SELECT BINTOHEX('AB') "bintohex" FROM DUMMY; bintohex 4142

BITAND


45 2012-12-19

Syntax

BITAND (n, m)

Description

Performs an AND operation on the bits of the arguments n and m. Both n and m must be non-negativeintegers. The BITAND function returns a result with BIGINT type.

Example

SELECT BITAND (255, 123) "bitand" FROM DUMMY; bitand 123

CEIL

Syntax

CEIL (n)

Description

Returns the first integer that is greater or equal to the value n.

Example

SELECT CEIL (14.5) "ceiling" FROM DUMMY; ceiling 15

COS

Syntax

COS (n)

Description

Returns the cosine of the angle, in radians, of the argument n.

Example

SELECT COS (0.0) "cos" FROM DUMMY; cos 1.0

COSH

Syntax

COSH (n)

Description

Computes the hyperbolic cosine of the argument n.

Example

SELECT COSH (0.5) "cosh" FROM DUMMY; cosh 1.1276259652063807


46 2012-12-19

COT

Syntax

COT (n)

Description

Computes the cotangent of a number n, where the argument is an angle expressed in radians.

Example

SELECT COT (40) "cot" FROM DUMMY; cot -0.8950829176379128

EXP

Syntax

EXP (n)

Description

Returns the result of the base of natural logarithms e raised to the power of the argument n.

Example

SELECT EXP (1.0) "exp" FROM DUMMY; exp 2.718281828459045

FLOOR

Syntax

FLOOR (n)

Description

Returns the largest integer not greater than the numeric argument n.

Example

SELECT FLOOR (14.5) "floor" FROM DUMMY; floor 14

GREATEST

Syntax

GREATEST (n1 [, n2]...)

Description

Returns the greatest value among the arguments: n1, n2, ...

Example

SELECT GREATEST ('aa', 'ab', 'ba', 'bb') "greatest" FROM DUMMY;


47 2012-12-19

greatest bb

HEXTOBIN

Syntax

HEXTOBIN (value)

Description

Converts a hexadecimal value to a binary value.

Example

SELECT HEXTOBIN ('1a') "hextobin" FROM DUMMY; hextobin 1A

LEAST

Syntax

LEAST (n1 [, n2]...)

Description

Returns the least value among the arguments: n1, n2...

Example

SELECT LEAST('aa', 'ab', 'ba', 'bb') "least" FROM DUMMY; least aa

LN

Syntax

LN (n)

Description

Returns the natural logarithm of the argument n.

Example

SELECT LN (9) "ln" FROM DUMMY; ln 2.1972245773362196

LOG

Syntax

LOG (b, n)

Description

Returns the natural logarithm of a number n base b. Base b must be a positive value greater than 1 and nmust be any positive value.

Example


48 2012-12-19

SELECT LOG (10, 2) "log" FROM DUMMY; log 0.30102999566398114

MOD

Syntax

MOD (n, d)

Description

Returns the remainder of a number n divided by a divisor d.

When n is negative this function acts differently to the standard computational modulo operation.

The following explains example of what the MOD function returns as the result.

If d is zero, then n is returned.If n is greater than 0 and n is less than d, then n is returned.If n is less than 0 and n is greater than d, then n is returned.In other case that those mentioned above, remainder of the absolute value of n divided by theabsolute value of d is used to calculate remainder. If n is less than 0, then the returned remainderfrom MOD is a negative number, and if n is greater than 0, then the returned remainder from MOD isa positive number.

Example

SELECT MOD (15, 4) "modulus" FROM DUMMY; modulus 3

SELECT MOD (-15, 4) "modulus" FROM DUMMY; modulus -3

POWER

Syntax

POWER (b, e)

Description

Calculates the base number b raised to the power of an exponent e.

Example

SELECT POWER (2, 10) "power" FROM DUMMY; power 1024.0

RAND

Syntax

DOUBLE RAND()

Description

Returns a pseudo-random value in the range of [0, 1.0). Its return value type is DOUBLE.

Example


49 2012-12-19

SELECT RAND() FROM DUMMY; RAND 3.907985046680551e-14

ROUND

Syntax

ROUND (n [, pos])

Description

Rounds argument n to the specified pos amount of places after the decimal point.

Example

SELECT ROUND (16.16, 1) "round" FROM DUMMY; round 16.2

SELECT ROUND (16.16, -1) "round" FROM DUMMY; round 20

SIGN

Syntax

SIGN (n)

Description

Returns the sign (positive or negative) of the numeric argument n. Returns 1 if n is a positive value,-1 if n is anegative value, and 0 if n is equal to zero.

Example

SELECT SIGN (-15) "sign" FROM DUMMY; sign -1

SIN

Syntax

SIN (n)

Description

Returns the sine of n, where the argument is an angle expressed in radians.

Example

SELECT SIN ( 3.141592653589793/2) "sine" FROM DUMMY; sine 1.0

SINH

Syntax

SINH (n)


50 2012-12-19

Description

Returns the hyperbolic sine of n, where the argument is an angle expressed in radians.

Example

SELECT SINH (0.0) "sinh" FROM DUMMY; sinh 0.0

SQRT

Syntax

SQRT (n)

Description

Returns the square root of the argument n.

Example

SELECT SQRT (2) "sqrt" FROM DUMMY; sqrt 1.4142135623730951

TAN

Syntax

TAN (n)

Description

Returns the tangent of n, where the argument is an angle expressed in radians.

Example

SELECT TAN (0.0) "tan" FROM DUMMY; tan 0.0

TANH

Syntax

TANH (n)

Description

Returns the hyperbolic tangent of the numeric argument n.

Example

SELECT TANH (1.0) "tanh" FROM DUMMY; tanh 0.7615941559557649

UMINUS

Syntax

UMINUS (n)


51 2012-12-19

UMINUS (n)

Description

Returns the negated value of the numeric argument n.

Example

SELECT UMINUS(-765) "uminus" FROM DUMMY;

uminus 765

String Functions

ASCII

Syntax

ASCII(c)

Description

Returns the integer ASCII value of the first byte in a string c.

Example

SELECT ASCII('Ant') "ascii" FROM DUMMY; ascii 65

CHAR

Syntax

CHAR (n)

Description

Returns the character with the ASCII value of a number n.

Example

SELECT CHAR (65) || CHAR (110) || CHAR (116) "character" FROM DUMMY; character Ant

CONCAT

Syntax

CONCAT (str1, str2)

Description

Returns a combined string consisting of str1 followed by str2. The concatenation operator (||) is identical tothis function.

Example

SELECT CONCAT ('C', 'at') "concat" FROM DUMMY; concat Cat


52 2012-12-19

LCASE

Syntax

LCASE(str)

Description

Converts all characters in string str to lowercase.

Note:The LCASE function is identical to the LOWER function.

Example

SELECT LCASE ('TesT') "lcase" FROM DUMMY; lcase test

LEFT

Syntax

LEFT (str, n)

Description

Returns the first n characters/bytes from the beginning of string str.

Example

SELECT LEFT ('Hello', 3) "left" FROM DUMMY; left Hel

LENGTH

Syntax

LENGTH(str)

Description

Returns the number of characters in string str.

Note:Supplementary plane Unicode characters, each of which occupies 6 bytes in CESU-8 encoding, arecounted as two characters.

Example

SELECT LENGTH ('length in char') "length" FROM DUMMY; length 14

LOCATE

Syntax

LOCATE (haystack, needle)

Description

Returns the position of a substring needle within a string haystack. Returns 0 if needle is not found within


53 2012-12-19

haystack. Returns NULL if haystack or needle is NULL, or if haystack is an empty string.

Example

SELECT LOCATE ('length in char', 'char') "locate" FROM DUMMY; Locate 11

SELECT LOCATE ('length in char', 'length') "locate" FROM DUMMY; Locate 1

SELECT LOCATE ('length in char', 'zchar') "locate" FROM DUMMY; Locate 0

LOWER

Syntax

LOWER(str)

Description

Converts all characters in string str to lowercase.

Note:The LOWER function is identical to the LCASE function.

Example

SELECT LOWER ('AnT') "lower" FROM DUMMY; lower ant

LPAD

Syntax

LPAD (str, n [, pattern])

Description

Pads the start of string str with spaces to make a string of n characters in length. If the pattern argument isprovided string str will be padded using sequences of these characters until the required length is met.

Example

SELECT LPAD ('end', 15, '12345') "lpad" FROM DUMMY; lpad 123451234512end

LTRIM

Syntax

LTRIM (str [, remove_set])

Description

Returns string str, trimmed of all leading spaces. If remove_set is specified, LTRIM removes all the characterscontained in this set from the start of string str. This process continues until a character not the inremove_set is reached.


54 2012-12-19

Note:remove_set is treated as a set of characters and not as a search string.

Example

SELECT LTRIM ('babababAabend','ab') "ltrim" FROM DUMMY; ltrim Aabend

NCHAR

Syntax

NCHAR (n)

Description

Returns the Unicode character with the integer code number n.

Example

SELECT NCHAR (65) "nchar" FROM DUMMY; nchar A

REPLACE

Syntax

REPLACE (original_string, search_string, replace_string)

Description

Searches in original_string for all occurrences of search_string and replaces them with replace_string.

If original_string is an empty string, then the result will be an empty string.If two overlapping substrings match the search_string in the original_string, then only the firstoccurrence will be replaced.If original_string does not contain any occurrence of search_string, then original_string will bereturned unchanged.If original_string, search_string, or replace_string are NULL then NULL is returned.

Example

SELECT REPLACE ('DOWNGRADE DOWNWARD','DOWN', 'UP') "replace" FROM DUMMY; replace UPGRADE UPWARD

RIGHT

Syntax

RIGHT(str, n)

Description

Returns the rightmost n characters/bytes of string str.

Example

SELECT RIGHT('HI0123456789', 3) "right" FROM DUMMY; right 789


55 2012-12-19

RPAD

Syntax

RPAD (str, n [, pattern])

Description

Pads the end of string str with spaces to make a string of n characters in length. If the pattern argument isprovided string str will be padded using sequences of these characters until the required length is met.

Example

SELECT RPAD ('end', 15, '12345') "right padded" FROM DUMMY; right padded end123451234512

RTRIM

Syntax

RTRIM (str [,remove_set ])

Description

Returns string str, trimmed of all trailing spaces. If remove_set is specified, RTRIM removes all the characterscontained in this set from the end of string str. This process continues until a character not the in remove_setis reached.

Note:remove_set is treated as a set of characters and not as a search string.

Example

SELECT RTRIM ('endabAabbabab','ab') "rtrim" FROM DUMMY; rtrim endabA

SUBSTR_AFTER

Syntax

SUBSTR_AFTER (str, pattern)

Description

Returns a substring of string str that follows the first occurrence of the pattern argument.

If str does not contain the pattern substring, then an empty string is returned.If pattern is an empty string, then str is returned.If str or pattern is NULL, then NULL is returned.

Example

SELECT SUBSTR_AFTER ('Hello My Friend','My ') "substr after" FROM DUMMY; substr after Friend

SUBSTR_BEFORE

Syntax

SUBSTR_BEFORE (str, pattern)


56 2012-12-19

Description

Returns a substring of string str before the first occurrence of the pattern argument in the target string.

If str does not contain the pattern substring, then an empty string is returned.If pattern is an empty string, then str is returned.If str or pattern is NULL, then NULL is returned.

Example

SELECT SUBSTR_BEFORE ('Hello My Friend','My') "substr before" FROM DUMMY; substr before Hello

SUBSTRING

Syntax

SUBSTRING (str, start_position [, string_length])

Description

Returns a substring of string str starting from start_position within the string. SUBSTRING can return theremaining part of a string from the start_position or optionally, a number of characters set by thestring_length parameter.

If start_position is less than 0, then it is considered as 1.If string_length is less than 1, then an empty string is returned.

Example

SELECT SUBSTRING ('1234567890',4,2) "substring" FROM DUMMY; substring 45

TRIM

Syntax

TRIM ([[LEADING | TRAILING | BOTH] trim_char FROM] str )

Description

Returns string str after removing leading and trailing spaces. The trimming operation is carried out eitherfrom the start (LEADING), end (TRAILING) or both(BOTH) ends of string str.

If either str or trim_char are a null values, then a NULL is returned.If no options are specified, TRIM removes both the leading and trailing substring trim_char from stringstr.If trim_char is not specified, then a single blank space is used.

Example

SELECT TRIM ('a' FROM 'aaa123456789aa') "trim both" FROM DUMMY; trim both 123456789 SELECT TRIM (LEADING 'a' FROM 'aaa123456789aa') "trim leading" FROM DUMMY; trim leading 123456789aa

UCASE

Syntax

UCASE (str)


57 2012-12-19

Description

Converts all characters in string str to uppercase.

Note:The UCASE function is identical to the UPPER function.

Example

SELECT UCASE ('Ant') "ucase" FROM DUMMY; ucase ANT

UNICODE

Syntax

UNICODE(c)

Description

Returns an integer containing the Unicode code point of the first character in the string, or NULL if the firstcharacter is not a valid encoding.

Example

SELECT UNICODE ('#') "unicode" FROM DUMMY; unicode 35

UPPER

Syntax

UPPER (str)

Description

Converts all characters in string str to uppercase.

Note:The UPPER function is identical to the UCASE function.

Example

SELECT UPPER ('Ant') "uppercase" FROM DUMMY; uppercase ANT

Window Functions

Miscellaneous Functions

COALESCE

Syntax

COALESCE (expression_list)

Description

Returns the first non-NULL expression from a list. At least two expressions must be contained inexpression_list, and all expressions must be comparable. The result will be NULL if all the arguments are


58 2012-12-19

NULL.

Example

CREATE TABLE coalesce_example (ID INT PRIMARY KEY, A REAL, B REAL); INSERT INTO coalesce_example VALUES(1, 100, 80); INSERT INTO coalesce_example VALUES(2, NULL, 63); INSERT INTO coalesce_example VALUES(3, NULL, NULL);

SELECT id, a, b, COALESCE (a, b*1.1, 50.0) "coalesce" FROM coalesce_example;

ID A B coalesce

1 100.0 80.0 100.0

2 NULL 63.0 69.30000305175781

3 NULL NULL 50.0

CURRENT_CONNECTION

Syntax

CURRENT_CONNECTION

Description

Returns the id of the current connection.

Example

SELECT CURRENT_CONNECTION "current connection" FROM DUMMY; current connection 2

CURRENT_SCHEMA

Syntax

CURRENT_SCHEMA

Description

Returns a string containing the current schema name.

Example

SELECT CURRENT_SCHEMA "current schema" FROM DUMMY; current schema SYSTEM

CURRENT_USER

Syntax

CURRENT_USER

Description

Returns the current user name at the current statement context. This will be user name which is currently atthe top of authorization stack.

Example

-- example showing basic function operation using SYSTEM user SELECT CURRENT_USER "current user" FROM DUMMY;


59 2012-12-19

SELECT CURRENT_USER "current user" FROM DUMMY; current user SYSTEM

-- definer-mode procedure declared by USER_A CREATE PROCEDURE USER_A.PROC1 LANGUAGE SQLSCRIPT SQL SECURITY DEFINER AS BEGIN SELECT CURRENT_USER "current user" FROM DUMMY; END; -- USER_B executing USER_A.PROC1 CALL USER_A.PROC1;

current user USER_A

-- invoker-mode procedure declared by USER_A CREATE PROCEDURE USER_A.PROC2 LANGUAGE SQLSCRIPT SQL SECURITY INVOKER AS BEGIN SELECT CURRENT_USER "current user" FROM DUMMY; END; -- USER_B is executing USER_A.PROC CALL USER_A.PROC2;

current user USER_B

GROUPING_ID

Syntax

GROUPING_ID(column_name_list)

Description

GROUPING_ID function can be used with GROUPING SETS to return multiple levels of aggregations in a singleresult set. GROUPING_ID returns an integer value to identify which grouping set each row belongs to. Eachcolumn in GROUPING_ID must be an element of the GROUPING SETS.

GROUPING_ID is assigned by converting the bit vector generated from GROUPING SETS to a decimal numberby treating the bit vector as a binary number. When a bit vector is composed, 0 is assigned to each columnspecified in the GROUPING SETS and 1 otherwise in the order it appears in the GROUPING SETS. By treatingthe bit vector as a binary number, this function returns an integer value as the output.

Example

SELECT customer, year, product, SUM(sales), GROUPING_ID(customer, year, product) FROM guided_navi_tab GROUP BY GROUPING SETS ( (customer, year, product), (customer, year), (customer, product), (year, product), (customer), (year), (product)); CUSTOMER YEAR PRODUCT SUM(SALES) GROUPING_ID(CUSTOMER,YEAR,PRODUCT) 1 C1 2009 P1 100 0 2 C1 2010 P1 50 0 3 C2 2009 P1 200 0 4 C2 2010 P1 100 0 5 C1 2009 P2 200 0 6 C1 2010 P2 150 0 7 C2 2009 P2 300 0 8 C2 2010 P2 150 0 9 C1 2009 a 300 1 10 C1 2010 a 200 1 11 C2 2009 a 500 1 12 C2 2010 a 250 1 13 C1 a P1 150 2 14 C2 a P1 300 2 15 C1 a P2 350 2 16 C2 a P2 450 2


60 2012-12-19

16 C2 a P2 450 2 17 a 2009 P1 300 4 18 a 2010 P1 150 4 19 a 2009 P2 500 4 20 a 2010 P2 300 4 21 C1 a a 500 3 22 C2 a a 750 3 23 a 2009 a 800 5 24 a 2010 a 450 5 25 a a P1 450 6 26 a a P2 800 6

IFNULL

Syntax

IFNULL (expression1, expression2)

Description

Returns the first not NULL input expression.

Returns expression1 if expression1 is not NULL.Returns expression2 if expression1 is NULL.Returns NULL if both input expressions are NULL.

Example

SELECT IFNULL ('diff', 'same') "ifnull" FROM DUMMY; ifnull diff

SELECT IFNULL (NULL, 'same') "ifnull" FROM DUMMY; ifnull same

SELECT IFNULL (NULL, NULL) "ifnull" FROM DUMMY; ifnull NULL

MAP

Syntax

MAP (expression, search1, result1 [, search2, result2] ... [, default_result])

Description

Searches for an expression within a set of search values and returns the corresponding result.

If the expression value is not found and default_result is defined, MAP returns default_result.If the expression value is not found and default_result is not defined, MAP returns NULL.

Note:Search values and corresponding results are always provided in search-result pairs.

Example

SELECT MAP(2, 0, 'Zero', 1, 'One', 2, 'Two', 3, 'Three', 'Default') "map" FROM DUMMY;

map Two

SELECT MAP(99, 0, 'Zero', 1, 'One', 2, 'Two', 3, 'Three', 'Default') "map" FROM DUMMY;


61 2012-12-19

map Default

SELECT MAP(99, 0, 'Zero', 1, 'One', 2, 'Two', 3, 'Three') "map" FROM DUMMY;

map NULL

NULLIF

Syntax

NULLIF (expression1, expression2)

Description

NULLIF compares the values of two input expressions. If the first expression equals the second expression,NULLIF returns NULL.

If expression1 does not equal expression2, NULLIF returns expression1.If expression2 is NULL, NULLIF returns expression1.

Example

SELECT NULLIF ('diff', 'same') "nullif" FROM DUMMY; nullif diff

SELECT NULLIF('same', 'same') "nullif" FROM DUMMY; nullif NULL

SESSION_CONTEXT

Syntax

SESSION_CONTEXT(session_variable)

Description

Returns the value of session_variable assigned to the current user.

The session_variable accessed can either be predefined or user-defined. Predefined session variables that canbe set by the client are 'APPLICATION', 'APPLICATIONUSER', and 'TRACEPROFILE'.

Session variables can be defined or modified using SET [SESSION] <variable_name> = <value> command, andunset using UNSET [SESSION] <variable_name>.

SESSION_CONTEXT returns an NVARCHAR with a maximum length of 512 characters.

Example

To read session variables use:

SELECT SESSION_CONTEXT('APPLICATION') "session context" FROM DUMMY;

session context HDBStudio

SESSION_USER

Syntax

SESSION_USER

Description


62 2012-12-19

Returns the user name of the current session.

Example

-- example showing basic function operation using SYSTEM user SELECT SESSION_USER "session user" FROM DUMMY; session user SYSTEM

-- definer-mode procedure declared by USER_A CREATE PROCEDURE USER_A.PROC1 LANGUAGE SQLSCRIPT SQL SECURITY DEFINER AS BEGIN SELECT SESSION_USER "session user" FROM DUMMY; END; -- USER_B is executing USER_A.PROC CALL USER_A.PROC1;

session user USER_B

-- invoker-mode procedure declared by USER_A CREATE PROCEDURE USER_A.PROC2 LANGUAGE SQLSCRIPT SQL SECURITY INVOKER AS BEGIN SELECT SESSION_USER "session user" FROM DUMMY; END; -- USER_B is executing USER_A.PROC CALL USER_A.PROC2;

session user USER_B

SYSUUID

SYSUUID

SYSUUID

Description

Returns a new universally unique identifier, generated by the connected SAP HANA instance. Each call ofSYSUUID returns a new UUID value. SYSUUID calls from multiple connections are internally serialized toguarantee unique value generation.

Example

SELECT SYSUUID FROM DUMMY; SYSUUID 4DE3CD576C79511BE10000000A3C2220


63 2012-12-19

SQL Statements

This chapter describes the SQL statements that are supported by the SAP HANA Database.

Data Definition StatementsData Manipulation StatementsSystem Management StatementsSession Management StatementsTransaction Management StatementsAccess Control StatementsData Import Export StatementsProcedural Statements

Data Definition Statements

ALTER AUDIT POLICY

Syntax

ALTER AUDIT POLICY <policy_name> <audit_mode>

Syntax Elements

<policy_name> ::= <identifier>

The name of the audit policy to be altered.

<audit_mode> ::= ENABLE | DISABLE

The audit mode enables or disables the audit policy.

ENABLE

Enables the audit policy.

DISABLE

Disables the audit policy.

Description

The ALTER AUDIT POLICY statement enables or disables an audit policy. <policy_name> must specify anexisting audit policy.

Only database users with the system privilege AUDIT ADMIN are allowed to alter an audit policy. Users withthis privilege can alter any audit policy, regardless of if they are the creator of the policy.

When an audit policy is created, it is created in the disabled state. Therefore the audit policy has to beenabled to make its audit actions take effect.

An audit policy can be disabled and enabled as often as required.

Configuration Parameters

The following configuration parameters for auditing are stored in global.ini, in the section auditingconfiguration:

global_auditing_state ( 'true' / 'false' )

Regardless of the number of enabled audit policies, auditing will only occur if the global_auditing_stateconfiguration parameter is set to true. The default is 'false'.

default_audit_trail_type ( 'SYSLOGPROTOCOL' / 'CSVTEXTFILE' )

Specifies how auditing results will be stored.

SYSLOGPROTOCOL: uses the system syslog.CSVTEXTFILE: stores audit information as comma-separated values in a text file.


64 2012-12-19

default_audit_trail_path

Specifies the file path where the CSVTEXTFILE should be stored.

The parameters above can be selected in the M_INIFILE_CONTENTS monitoring view if the user has therequired system privileges. The parameters can only be observed in a monitoring view when they have beenexplicitly defined.

System Tables and Monitoring Views

AUDIT_POLICY: shows all audit policies and their states. M_INIFILE_CONTENTS: shows the database system configuration parameters.

Only users with system privilege CATALOG READ, DATA ADMIN or INIFILE ADMIN can view the content ofthe M_INIFILE_CONTENTS view. For all other database users this view will be empty.

Example

For this example you need to first create an audit policy called priv_audit using the following statement:

CREATE AUDIT POLICY priv_audit AUDITING SUCCESSFUL GRANT PRIVILEGE, REVOKE PRIVILEGE, GRANT ROLE, REVOKE ROLE LEVEL CRITICAL;

Now you can enable this audit policy with:

ALTER AUDIT POLICY priv_audit ENABLE;

You can also disable the same audit policy with:

ALTER AUDIT POLICY priv_audit DISABLE;

ALTER FULLTEXT INDEX

Syntax

ALTER FULLTEXT INDEX <index_name> <alter_fulltext_index_option>

Syntax Elements

<index_name> ::= <identifier>

The identifier of the fulltext index to be altered.

<alter_fulltext_index_option> ::= <fulltext_parameter_list> | <queue_command> QUEUE

Specifies whether the parameters of the fulltext index or the state of the fulltext index queue should bechanged. The latter is only possible for an asynchronous explicit fulltext index.

<fulltext_parameter_list> ::= <fulltext_parameter> [{, <fulltext_parameter>} ...]

A list of fulltext index parameters that are to be changed.

<fulltext_parameter> ::= FUZZY SEARCH INDEX <on_off> | PHRASE INDEX RATIO <index_ratio>

Fulltext index parameters.

FUZZY SEARCH INDEX <on_off> <on_off> ::= ON | OFF

Defines if a fuzzy search index will be used. ON switches on the fuzzy seach index, and OFF turns the fuzzysearch index off.

PHRASE INDEX RATIO <index_ratio>

Specifies the phrase index ratio.


65 2012-12-19

<index_ratio> ::= <exact_numeric_literal>

Specifies the percentage of the phrase index ratio, the value used must be between 0.0 and 1.0

<queue_command> ::= FLUSH | SUSPEND | ACTIVATE

Specifies an action to be performed on the index queue.

FLUSH

Updates the fulltext index with the documents in the queue that have already been processed.

SUSPEND

Suspends the fulltext index processing queue.

ACTIVATE

Activates the fulltext index processing queue.

Description

The ALTER FULLTEXT INDEX command either change the parameters of a fulltext index or the state of anindex processing queue. The queue is a mechanism used to enable a fulltext index to operate in asynchronousmanner, i.e. inserts will not block until a document is processed.

Examples

You create table T and then add two indices i1 and i2. Index i1 is a synchronous fulltext index, i2 is anasynchronous fulltext index.

CREATE COLUMN TABLE T (c1 NVARCHAR(1000), c2 NVARCHAR(1000)); CREATE FULLTEXT INDEX i1 ON T (c1) SYNC; CREATE FULLTEXT INDEX i2 ON T (c2) ASYNC;

You alter the phrase index to 30 percent and turn on the fuzzy search index for indexes i1 and i2.

ALTER FULLTEXT INDEX i1 PHRASE INDEX RATIO 0.3 FUZZY SEARCH INDEX ON; ALTER FULLTEXT INDEX i2 PHRASE INDEX RATIO 0.3 FUZZY SEARCH INDEX ON;

You suspend the queue for fulltext index i2.

ALTER FULLTEXT INDEX i2 SUSPEND QUEUE;

You update the fulltext index with the documents in the queue that have already been processed.

ALTER FULLTEXT INDEX i2 FLUSH QUEUE;

ALTER INDEX

Syntax

ALTER INDEX <index_name> REBUILD

Syntax Elements

<index_name>::= <identifier>

Specifies the name of the index to be rebuilt.

Description

The ALTER INDEX statement rebuilds an index.

Example

You create table A and an index i on column b of table A.


66 2012-12-19

CREATE TABLE A (a INT, b NVARCHAR(10), c NVARCHAR(20)); CREATE INDEX i ON A(b);

You rebuild index i.

ALTER INDEX i REBUILD;

ALTER SEQUENCE

Syntax

ALTER SEQUENCE <sequence_name> [<restart_with>] [<parameter_list>] [RESET BY <reset_by_subquery>]

Syntax Elements

<sequence_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The name of the sequence to be altered with optional schema name.

<sequence_parameter_restart_with> ::= RESTART WITH <restart_value>

The starting value of the sequence. If you do not specify a value for the RESTART WITH clause, the currentvalue of the sequence is used.

<restart_value> ::= <unsigned_integer>

An integer value between 0 and 4611686018427387903 which will be the first value provided by thesequence generator.

<parameter_list> ::= <sequence_parameter>[{, <sequence_parameter>}...]

Defines the paramter list.

<sequence_parameter>::= INCREMENT BY <increment_value> | MAXVALUE <maximum_value> | NO MAXVALUE | MINVALUE <minimum_value> | NO MINVALUE | CYCLE | NO CYCLE

A set of parameters that can be used with ALTER SEQUENCE. For parameter details see below.

INCREMENT BY <increment_value>

The sequence increment.

<increment_value> ::= <signed_integer>

An integer value which will be used to increment or decrement the sequence value.

MAXVALUE <maximum_value>

Defines the maximum value that can be generated by the sequence.

<maximum_value> ::= <unsigned_integer>

A positive integer which defines the maximum value generated by the sequence. The maximum value mustbe between 0 and 4611686018427387903.

NO MAXVALUE

When the NO MAXVALUE directive is used the maximum value for an ascending sequence will be4611686018427387903 and the maximum value for a descending sequences will be -1.

MINVALUE <minimum_value>


67 2012-12-19

Defines the minimum value that can be generated by the sequence.

<minimum_value> ::= <unsigned_integer>

A positive integer which defines the minimum value generated by the sequence. The minimum value must bebetween 0 and 4611686018427387903.

NO MINVALUE

When the NO MINVALUE directive is used, the minimum value for an ascending sequence is 1 and theminimum value for a descending is -4611686018427387903.

CYCLE

When the CYCLE directive is used the sequence number will be restarted after it reaches its maximum orminimum value.

NO CYCLE

When the NO CYCLE directive is used the sequence number will not be restarted after it reaches its maximumor minimum value.

<reset_by_subquery> ::= <subquery>

During a restart of the database, the system automatically executes the RESET BY statement and thesequence value is restarted with the value determined from the RESET BY subquery.

For details about subqueries please see Subquery

Description

The ALTER SEQUENCE statement is used to alter the parameters of a sequence generator.

Examples

Example 1

You create table A and a sequence seq. Sequence seq, when reset, will start from the value of the selectstatement shown.

CREATE TABLE A (a INT); CREATE SEQUENCE seq RESET BY SELECT IFNULL(MAX(a), 0) + 1 FROM A;

You change the starting sequence value of sequence seq to 2.

ALTER SEQUENCE seq RESTART WITH 2;

You change the maximum value of sequence s to 100, and specify that it does not have a minimum value.

ALTER SEQUENCE seq MAXVALUE 100 NO MINVALUE;

You change the incremental value of sequence seq to 3, and specify that the sequence will not restart uponreaching its maximum or minimum value.

ALTER SEQUENCE seq INCREMENT BY 3 NO CYCLE;

Example 2

You create table B, with column a. You create a sequence s1 with a reset-by subquery based on table B.

CREATE TABLE B (a INT); CREATE SEQUENCE s1 RESET BY SELECT IFNULL(MAX(a), 0) + 1 FROM B;

You change the reset-by subquery of sequence s1 to the maximum value contained in column a of table B.

ALTER SEQUENCE s1 RESET BY SELECT MAX(a) FROM B;

ALTER TABLE


68 2012-12-19

Syntax

ALTER TABLE <table_name> [<add_column_clause>] [<drop_column_clause>] [<alter_column_clause>] [<add_constraint_clause>] [<drop_primary_key_clause>][<drop_constraint_clause>] [<preload_clause>] [<table_conversion_clause>] [<move_clause>] [<add_range_partition_clause>] [<drop_range_partition_clause>] [<partition_clause>] [<persistent_merge_option>] [<delta_log_option>] [<auto_merge_option>] [<unload_priority>] [<schema_flexibility_option>]

Syntax elements:<table_name><add_column_clause><drop_column_clause><alter_column_clause><add_constraint_clause><drop_primary_key_clause><drop_constraint_clause><preload_clause><table_conversion_clause><add_range_partition_clause><drop_range_partition_clause><partition_clause><persistent_merge_option><delta_log_option><auto_merge_option><unload_priority><schema_flexibility_option>

Syntax Elements

<table_name>

<table_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The identifier of the table to be altered, with optional schema name.

<add_column_clause>

<add_column_clause> ::= ADD ( {<column_definition> [<column_constraint>]} [{, <column_definition> [<column_constraint>]}...] )

Adds one or more columns to the specified table.

<drop_column_clause>

<drop_column_clause> ::= DROP ( <column_name>[{, <column_name>} ...] )

Removes one or more columns from the specified table.

<alter_column_clause>

<alter_column_clause> ::= ALTER ( <column_definition> [<column_constraint>] [{, <column_definition> [<column_constraint>]}...] )

Alters one or more column definitions.Restrictions:

To prevent data loss, only increasing the size of a column data type is allowed. For example, changingfrom nvarchar(20) to nvarchar(10) or from integer to tinyint will raise an error.ALTER does not currently follow data type conversion rules.Adding NOT NULL constraint to an existing column is allowed if either of the following cases are true:

The table is empty.The default value is specified when the table contains data.

<column_definition> ::= <column_name> {<data_type> | <lob_data_type>} [<column_store_data_type>] [<ddic_data_type>] [DEFAULT <default_value>]


69 2012-12-19

[GENERATED ALWAYS AS <expression>] [<schema_flexibility>] [<fuzzy_search_index>] [<fuzzy_search_mode>]

Defines a table column.

<column_name> ::= <identifier>

The table column name.

<data_type> ::= DATE | TIME | SECONDDATE | TIMESTAMP | TINYINT | SMALLINT | INTEGER | BIGINT | SMALLDECIMAL | DECIMAL | REAL | DOUBLE | VARCHAR | NVARCHAR | ALPHANUM | SHORTTEXT | VARBINARY | TEXT

<lob_data_type> ::= <lob_type_name> [MEMORY THRESHOLD <memory_threshold_value>] <lob_type_name> ::= BLOB | CLOB | NCLOB

<column_store_data_type> ::= CS_ALPHANUM | CS_INT | CS_FIXED | CS_FLOAT | CS_DOUBLE | CS_DECIMAL_FLOAT | CS_FIXED(p-s, s) | CS_SDFLOAT | CS_STRING | CS_UNITEDECFLOAT | CS_DATE | CS_TIME | CS_FIXEDSTRING | CS_RAW | CS_DAYDATE | CS_SECONDTIME | CS_LONGDATE | CS_SECONDDATE

<ddic_data_type> ::= DDIC_ACCP | DDIC_ALNM | DDIC_CHAR | DDIC_CDAY | DDIC_CLNT | DDIC_CUKY | DDIC_CURR | DDIC_D16D | DDIC_D34D | DDIC_D16R | DDIC_D34R | DDIC_D16S | DDIC_D34S | DDIC_DATS | DDIC_DAY | DDIC_DEC | DDIC_FLTP | DDIC_GUID | DDIC_INT1 | DDIC_INT2 | DDIC_INT4 | DDIC_INT8 | DDIC_LANG | DDIC_LCHR | DDIC_MIN | DDIC_MON | DDIC_LRAW | DDIC_NUMC | DDIC_PREC | DDIC_QUAN | DDIC_RAW | DDIC_RSTR | DDIC_SEC | DDIC_SRST | DDIC_SSTR | DDIC_STRG | DDIC_STXT | DDIC_TIMS | DDIC_UNIT | DDIC_UTCM | DDIC_UTCL | DDIC_UTCS | DDIC_TEXT | DDIC_VARC | DDIC_WEEK

The available data types. Please see Data Types.

<memory_threshold_value> ::= <unsigned_integer> | NULL

Controls if LOB data should be stored in memory or not according to the following conditions:

If <memory_threshold_value> is not provided, lob is stored in memory by default.If <memory_threshold_value> is provided and its LOB size is bigger than memory threshold value, LOBdata is stored on disk.If <memory_threshold_value> is provided and its LOB size is equal or less than memory thresholdvalue, LOB data is stored in memory.If <memory_threshold_value> is NULL, all LOB data is stored in memory.If <memory_threshold_value> is 0, all LOB data is stored in disk.

DEFAULT <default_value> ::= NULL | <string_literal> | <signed_numeric_literal> | <unsigned_numeric_literal>

Specifies a value to be assigned to the column if an INSERT statement does not provide a value for the column

GENERATED ALWAYS AS <expression>

Specifies the expression to generate the column value in runtime.

<schema_flexibility> ::= [ENABLE | DISABLE] SCHEMA FLEXIBILITY

Specifies the column is dynamic:

ENABLE produces a dynamic column.DISABLE (default value) produces a static column.

<fuzzy_search_index> ::= FUZZY SEARCH INDEX [ON | OFF]

Turns a fuzzy search index on or off. OFF is the default.

<fuzzy_search_mode> ::= FUZZY SEARCH MODE [<string_literal> | NULL]

Sets the fuzzy search mode with the value of <string_literal>. If NULL is specified, the fuzzy search mode is


70 2012-12-19

reset.

<column_constraint> ::= NULL | NOT NULL | <unique_specification> | <references_specification>

The column constraint rules.For NULL, NOT NULL and <unique_specification>, see below. For <references_specification>, see References Specification

NULL

If NULL is specified it is not considered a constraint, it represents that a column that may contain a null value.The default is NULL.

NOT NULL

The NOT NULL constraint prohibits a column value from being NULL.

<unique_specification> ::= UNIQUE [<unique_tree_type_index>] | PRIMARY KEY [<unique_tree_type_index>]

Specifies unique constraints. If the index type is omitted, the SAP HANA database chooses the appropriateindex by considering the column data type. If the index type is not specified, the SAP HANA database willautomatically select an index type as follows:

Index type Criteria

CPBTREE

- character s tring types . - binary s tring types .- decimal types .- when the constra int i s a compos i te key.- when the constra int i s a non-unique constra int

BTREE Al l other cases than speci fied for CPBTREE

UNIQUE

Specifies a column as a unique key.A composite unique key enables the specification of multiple columns as a unique key. With a uniqueconstraint, multiple rows cannot have the same value in the same column.

PRIMARY KEY

A primary key constraint is a combination of a NOT NULL constraint and a UNIQUE constraint. It prohibitsmultiple rows from having the same value in the same column.

<unique_tree_type_index> ::= BTREE | CPBTREE;

Specifies the index type.

BTREE

BTREE specifies a B+-tree index. B+-tree is a tree that maintains sorted data which performs efficientinsertion, deletion and search of records.

CPBTREE

CPBTREE specifies a CPB+-tree index. CPB+-tree stands for Compressed Prefix B+-Tree, which is based on pkB-tree. CPB+-tree is a very small index because it uses 'partial key' that is only part of full key in index nodes.CPB+-tree shows better performance than B+-Tree for larger keys.

<add_constraint_clause>

<add_constraint_clause> ::= ADD [CONSTRAINT <constraint_name>] <table_constraint> <constraint_name> ::= <identifier>


71 2012-12-19

Adds a table constraint.

<table_constraint> ::= <unique_constraint_definition> | <referential_constraint_definition>

The table constraint can be either a unique constraint or a referential constraint.

<unique_constraint_definition> ::= <unique_specification> (<unique_column_name_list>)

The unique specification see Unique specifcation

<unique_column_name_list> ::= <unique_column_name>[{, <unique_column_name>}...]

Speicfies the unique column name list which can have one or more column names.

<unique_column_name> ::= <identifier>

A column name identifier.

<referential_constraint_definition> ::= FOREIGN KEY (<referencing_column_name_list>) <references_specification>

Specifies a referential constraint.

<referencing_column_name_list> ::= <referencing_column_name>[{, <referencing_column_name>}...]

Specifies the referencing column name list which can have one or more column names.

<referencing_column_name> ::= <identifier>

The identifier of a referencing column.

<references_specification> ::= REFERENCES <referenced_table> [(<referenced_column_name_list>)] [<referential_triggered_action>]

Specifies the referenced table, with optional column name list and trigger action. If<referenced_column_name_list> is specified, there will be one-to-one correspondence between<column_name> of <column_definition> (see column definition) and <referenced_column_name>. If it is notspecified, there will be one-to-one correspondence between <column_name> of <column_definition> and thecolumn name of the referenced table's primary key.

<referenced_column_name_list> ::= <referenced_column_name>[{, <referenced_column_name>}...]

Specifies the referenced column name list, which can have one or more column names.

<referenced_table> ::= <identifier>

The identifier of a table to be referenced.

<referenced_column_name> ::= <identifier>

The identifier of the column name to be referenced.

<referential_triggered_action> ::= <update_rule> [<delete_rule>] | <delete_rule> [<update_rule>]

Specifies an update rule with optional delete rule or a delete rule with optional update rule. The order inwhich they are provided provides an order of precedence for execution.

<update_rule> ::= ON UPDATE <referential_action> <referential_action> ::= CASCADE | RESTRICT | SET DEFAULT | SET NULL

The following UPDATE referential actions are possible:


72 2012-12-19

ActionName

Update Action

RESTRICTAny updates to a referenced table are prohibi ted i f there are any matched records in thereferencing table. This i s the defaul t action.

CASCADEIf a record i s updated in the referenced table, the corres ponding records in the referencing tableare a lso updated with the same va lues .

SET NULLIf a record i s updated in the referenced table, the corres ponding records in the referencing tableare a lso updated with nul l va lues .

SETDEFAULT

If a record i s updated in the referenced table, the corres ponding records in the referencing tableare a lso updated with thei r defaul t va lues

<delete_rule> ::= ON DELETE <referential_action>

The following DELETE referential actions are possible:

ActionName

Delete Action

RESTRICTAny deletions to a referenced table are prohibi ted i f there are any matched records in thereferencing table. This i s the defaul t action.

CASCADEIf a record in the referenced table i s deleted, the corres ponding records in the referencing tableare a lso deleted.

SET NULLIf a record in the referenced table i s deleted, the corres ponding records in the referencing tableare s et to nul l .

SETDEFAULT

If a record in the referenced table i s deleted, the corres ponding records in the referencing tableare s et to thei r defaul t va lues .

<drop_primary_key_clause>

<drop_primary_key_clause> ::= DROP PRIMARY KEY

Drops the primary key constraint.

<drop_constraint_clause>

<drop_constraint_clause> ::= DROP CONSTRAINT <constraint_name> <constraint_name> ::= <identifier>

Drops a unique or referential constraint.

<preload_clause>

<preload_clause> ::= PRELOAD ALL | PRELOAD ( <column_name> ) | PRELOAD NONE

Sets or removes the preload flag of the given tables or columns.When the preload flag is set tables are automatically loaded into memory after an index server start. Thecurrent status of the preload flag is visible in the system table TABLES in the PRELOAD column. Possiblevalues are 'FULL', 'PARTIALLY' and 'NO'. Also in system table TABLE_COLUMNS in column PRELOAD withpossible values being 'TRUE' or 'FALSE'.

PRELOAD ALL

Sets preload flags of all columns in the table.

PRELOAD ( <column_name> )

Sets the flags of the specified column.

PRELOAD NONE

Removes the preload flag from all columns.

<table_conversion_clause>


73 2012-12-19

<table_conversion_clause> ::= [ALTER TYPE] {ROW [THREADS <number_of_threads>] | COLUMN [THREADS <number_of_threads> [BATCH <batch_size>]]}

Converts the table storage from ROW to COLUMN or from COLUMN to ROW.

ROW

Converts the table to ROW storage.

COLUMN

Converts the table to COUMN storage.

THREADS <number_of_threads> <number_of_threads> ::= <unsigned_integer>

Specifies how many parallel execution threads should be used for the table conversion. The optimal value forthe number of threads is the number of available CPU cores. If THREADS is not provided the default value ofthe number of CPU cores specified in the indexserver.ini file will be used.

BATCH <batch_size> <batch_size> ::= <unsigned_integer>

Specifies the number of rows to be inserted in a batch. If BATCH is not specified the default value of2,000,000 will be used. Inserts into column tables will be immediately committed after every <batch_size>records have been inserted. BATCH option can be used only when a table is converted from ROW to COLUMNstorage.

<move_clause>

<move_clause> ::= MOVE [PARTITION <partition_number>] TO <indexserver_host_port> [PHYSICAL] | MOVE [PARTITION <partition_number>] PHYSICAL

Moves a table to another location in a distributed environment.

PARTITION <partition_number> <partition_number> ::= <unsigned_integer>

For partitioned tables, specifies the partition to be moved. If you attempt to move a partitioned tablewithout specifying a <partition_number> an error is returned.

<indexserver_host_port> ::= <unsigned_integer>

The internal indexserver port number where the table is to be moved.

PHYSICAL

Specifies that a column store tables persistence storage is moved immediately to the target host. If thePHYSICAL option is not specified the table move will create a link inside the new host persistence pointing tothe old host persistence. The link will be removed on the next merge or upon execution of another moveoperation not using the TO <indexserver_host_port> clause.

The PHYSICAL keyword is only for column store tables. Row store tables are always moved immediately.

<add_range_partition_clause>

<add_range_partition_clause> ::= ADD <range_partition_clause>

Adds a partition for tables partitioned with RANGE, HASH RANGE, ROUNDROBIN RANGE.

<range_partition_clause> ::= {<from_to_spec> | <single_spec>} [, PARTITION OTHERS]

The range specifier for a new partition.For syntax details please see partition range specifier below.

<drop_range_partition_clause>


74 2012-12-19

<drop_range_partition_clause> ::= DROP <range_partition_clause>

Drops a partition for tables partitioned with RANGE, HASH RANGE, ROUNDROBIN RANGE.

<partition_clause>

<partition_clause> ::= PARTITION BY <hash_partition> [, <range_partition> | , <hash_partition>] | PARTITION BY <range_partition> | PARTITION BY <roundrobin_partition> [,<range_partition>]

The PARTITION BY option partitions a table using the selected rules. For more information about partitioning please see the "Table Partitioning in the SAP HANA database"section in the "SAP HANA Administration guide" available from the SAP HANA Appliance page.

<hash_partition> ::= HASH (<partition_expression> [{<partition_expression>,}...]) PARTITIONS {<num_partitions> | GET_NUM_SERVERS()}

Partitions the created table using a hash partitioning scheme.

<range_partition> ::= RANGE (<partition_expression>) (<range_spec>, ...)

Partitions the created table using a range partitioning scheme.

<roundrobin_partition> ::= ROUNDROBIN PARTITIONS {<num_partitions> | GET_NUM_SERVERS()} [, <range_partition>]

Partitions the created table using a round robin partitioning scheme.

GET_NUM_SERVERS()

Returns the number of servers available in the partitioning cluster.

<range_spec> ::= {<from_to_spec> | <single_spec>} [{{<from_to_spec> | <single_spec>},} ...] [, PARTITION OTHERS]

The range specifier for a partition.

<from_to_spec> ::= PARTITION <lower_value> <= VALUES < <upper_value>

Specifies a partition using lower and upper values of a <partition_expression>.

<single_spec> ::= PARTITION VALUE = <target_value>

Specifies a partition using a single value of a <partition_expression>.

PARTITION OTHERS

Specifies that all other values that are not covered by the partition specification will be gathered into onepartition.

<partition_expression> ::= <column_name> | YEAR(<column_name>) | MONTH(<column_name>)

The specifier used to segregate data into partitions.

<lower_value> ::= <string_literal> | <numeric_literal>

The lower value of a partition specifier.

<upper_value> ::= <string_literal> | <numeric_literal>

The upper value of a partition specifier.

<target_value> ::= <string_literal> | <numeric_literal>

The target value of a single partition specifier.

<num_partitions> ::= <unsigned_integer>


75 2012-12-19

http://help.sap.com/hana_appliance

The number of partitions to be created for the table.

<merge_partition_clause> ::= MERGE PARTITIONS

Merges all parts of a partitioned table into a non-partitioned table.

<persistent_merge_option>

<persistent_merge_option> ::= {ENABLE | DISABLE} PERSISTENT MERGE

Enables or disables persistent merging.

When enabled the merge-manager uses persistent merges for the given table. This is the defaultbehaviour.When disabled the merge-manager uses main-memory merges instead of persistent merges for thegiven table.

<delta_log_option>

<delta_log_option> ::= {ENABLE | DISABLE} DELTA LOG

Enables or disables delta logging for table.

After enabling, you have to perform a savepoint to be certain that all data is persisted. Also you have toperform a data backup, otherwise it will not be possible to recover this data.

If logging is disabled, log entries will not be persisted for this table. Changes to this table will only be writtento the data store when a savepoint is carried out. This can cause loss of committed transaction should theindexserver terminate. In the case of a termination, you have to truncate this table and insert all data again.

Note:You should only use this option during initial load.

<auto_merge_option>

<auto_merge_option> ::= {ENABLE | DISABLE} AUTOMERGE

Enables or disables automatic delta merge on the specified table.

<unload_priority>

<unload_priority_option> ::= UNLOAD PRIORITY <unload_priority>

UNLOAD PRIORITY specifies that priority of table to be unloaded from memory.

<unload_priority> ::= <digit>

Sets the priority of table to be unloaded from memory. It can be 0 ~ 9, where 0 means not-unloadable and 9means earliest unload.

<schema_flexibility_option>

<schema_flexibility_option> ::= {ENABLE | DISABLE} SCHEMA FLEXIBILITY

Enables or disables schema flexibility for the specified table.

Description

The ALTER TABLE statement changes the definition of a table.

Examples

You create Table t, and then alter default value of column b to 10.

CREATE TABLE t (a INT, b INT); ALTER TABLE t ALTER (b INT DEFAULT 10);

You alter table t adding a new column c.


76 2012-12-19

ALTER TABLE t ADD (c NVARCHAR(10) DEFAULT 'NCHAR');

You create a primary key constraint, prim_key, on columns a and b of table t.

ALTER TABLE t ADD CONSTRAINT prim_key PRIMARY KEY (a, b);

You change the table type of table t to COLUMN storage.

ALTER TABLE t COLUMN;

You set the preload flags of column b and c on table t.

ALTER TABLE t PRELOAD (b, c);

You partition table t with a RANGE partition, and then add an additional partition.

ALTER TABLE t PARTITION BY RANGE (a) (PARTITION VALUE = 1, PARTITION OTHERS); ALTER TABLE t ADD PARTITION 2 <= VALUES < 10;

You disable delta logging of table t.

ALTER TABLE t DISABLE DELTA LOG;

You change the unload priority of table t to 2:

ALTER TABLE t UNLOAD PRIORITY 2;

You create table R. Then you alter table R adding a unique constraint UK.

CREATE TABLE R (A INT PRIMARY KEY, B NVARCHAR(10)); ALTER TABLE R ADD CONSTRAINT UK UNIQUE (B);

You drop the unique constraint UK from table R.

ALTER TABLE R DROP CONSTRAINT UK;

You create table S. You add a referential constraint FK to table S that references column A of table R withdelete cascade option.

CREATE TABLE S (FA INT, B NVARCHAR(10)); ALTER TABLE S ADD CONSTRAINT FK FOREIGN KEY(FA) REFERENCES R(A) ON DELETE CASCADE;

You create table T1 and then turn on the schema flexibility of the table. You test the schema flexibility oftable T1 by inserting a new record. A new column F is automatically created and a record is inserted:

CREATE COLUMN TABLE T1 (C INT); ALTER TABLE T1 ENABLE SCHEMA FLEXIBILITY; INSERT INTO T1 (C, F) VALUES (1, 'NVAR1');

You create table T2 with a fuzzy search index and fuzzy search mode. You then switch off the fuzzy searchindex of COL1 and finally reset fuzzy search mode of COL2.

CREATE COLUMN TABLE T2 (KEY INT, COL1 VARCHAR(10) FUZZY SEARCH INDEX ON, COL2 NVARCHAR(10) FUZZY SEARCH MODE 'postcode'); ALTER TABLE T2 ALTER (COL1 VARCHAR(10) FUZZY SEARCH INDEX OFF); ALTER TABLE T2 ALTER (COL2 NVARCHAR(10) FUZZY SEARCH MODE NULL);

CREATE AUDIT POLICY

Syntax

CREATE AUDIT POLICY <policy_name> AUDITING <audit_status_clause> <audit_actions> LEVEL <audit_level>

Syntax Elements



77 2012-12-19

The name of the audit policy to be created.

<audit_status_clause> ::= SUCCESSFUL | UNSUCCESSFUL | ALL

Defines whether successful, unsuccessful or all executions of the specified audit actions will be audited.

<audit_actions> ::= <actions_for_user_clause> | <audit_action_list_clause> | <target_audit_action_list_clause>

The audit actions for the audit policy.

<user_name> ::= <identifier>

The username of the user to be audited by the audit policy.

<actions_for_user_clause> ::= ACTIONS FOR <user_name>[{, <user_name>}...]

Audits actions for a user or a set of users.

<audit_action_list_clause> ::= <audit_action_list> [FOR <user_name>[{, <user_name>}...] <audit_action_list> ::= <audit_action_name>[{, <audit_action_name>}...] <audit_action_name> ::= GRANT PRIVILEGE | REVOKE PRIVILEGE | GRANT STRUCTURED PRIVILEGE | REVOKE STRUCTURED PRIVILEGE | GRANT APPLICATION PRIVILEGE | REVOKE APPLICATION PRIVILEGE | GRANT ROLE | REVOKE ROLE | GRANT ANY | REVOKE ANY | CREATE USER | DROP USER | CREATE ROLE | DROP ROLE | ENABLE AUDIT POLICY | DISABLE AUDIT POLICY | CREATE STRUCTURED PRIVILEGE | DROP STRUCTURED PRIVILEGE | ALTER STRUCTURED PRIVILEGE | CONNECT | SYSTEM CONFIGURATION CHANGE | SET SYSTEM LICENSE | UNSET SYSTEM LICENSE | ALTER USER | REPOSITORY_ACTIVATE | DROP TABLE

Audits specific system actions, optionally limited to a user or a set of users. For more information on auditactions see Audit Action Table.

<target_audit_action_list> ::= <target_audit_action_list> [FOR <user_name>[{, <user_name>}...] <target_audit_action_list> ::= <target_audit_action_name>[{, <target_audit_action_name>}...] ON <object_name>[{, <object_name>}...] <target_audit_action_name> ::= INSERT | UPDATE | DELETE | SELECT | EXECUTE

Audits actions on a database object or set of objects. Optionally this auditing can be limited to a user or a setof users. Only objects of type table, view, and procedure can be specified in the <target_audit_action_list>.Synonyms and sequences cannot be selected as objects for audit policies. Furthermore only these<target_audit_action_name>s can be combined with an object. The following table shows an overview ofauditable actions on objects.

Action Table View Procedure

DELETE YES YES ---

INSERT YES YES ---

SELECT YES YES ---

UPDATE YES YES ---

EXECUTE --- --- YES

<object_name> ::= <table_name> | <view_name> | <procedure_name>

<table_name> ::= [<schema_name>.]<identifier> <view_name> ::= [<schema_name>.]<identifier> <procedure_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

Specifies a database object for the target audit action.


78 2012-12-19

<audit_level> ::= EMERGENCY | ALERT | CRITICAL | WARNING | INFO

Assigns an audit policy to an audit level. Possible levels, in decreasing order of importance, are:

EMERGENCY.ALERT.CRITICAL.WARNING.INFO.

Description

The CREATE AUDIT POLICY statement creates a new audit policy. This audit policy can then be enabled andwill cause the auditing of the specified audit actions to occur. Only database users having the system privilege AUDIT ADMIN are allowed to create an audit policy. The specified audit policy name must be unique not match the name of an existing audit policy. An audit policy defines which audit actions will be audited. Audit policies need to be enabled for auditing tooccur. One audit policy can contain one of the following:

non-restricted auditing for n (>=1) usersauditing for actions not restricted to objectsauditing for actions which are restricted to objects.

For the last two alternatives listed, an optional restriction for user(s) is available.

For auditing to occur, audit policies have to be created and enabled. Also the configuration parameterglobal_auditing_state (see below) has to be set to true.

Audit Actions

The table below contains the available audit actions. Each of the audit actions are in a specific group, auditactions in the same group can be combined into one audit policy.


79 2012-12-19

Audit Action NameGroupNumber

Audit Operation Comment

GRANT PRIVILEGE 1 granting of privi leges to users or roles

REVOKE PRIVILEGE 1 revoking of privi leges from users or roles

GRANT STRUCTUREDPRIVILEGE

1granting of s tructured/ana lytica l privi leges tousers or roles

REVOKE STRUCTUREDPRIVILEGE

1revoking of s tructured/ana lytica l privi legesfrom users or roles

GRANT APPLICATIONPRIVILEGE

1granting of appl ication privi leges to users orroles

REVOKE APPLICATIONPRIVILEGE

1revoking of appl ication privi leges from usersor roles

GRANT ROLE 1 granting of roles to users or roles

REVOKE ROLE 1 revoking of roles from users or roles

GRANT ANY 1granting of privi leges , s tructured privi leges orroles to users or roles

REVOKE ANY 1revoking of privi leges , s tructured privi leges orroles from users or roles

CREATE USER 2 creation of users

DROP USER 2 dropping of users

ALTER USER 2 a l tering of users

CREATE ROLE 2 creation of roles

DROP ROLE 2 dropping of roles

CONNECT 3 creation of a user connection to the database

SYSTEMCONFIGURATIONCHANGE

4changes to the sys tem configuration (e.g.INIFILE)

ENABLE AUDIT POLICY 5 activation of audi t pol icies

DISABLE AUDITPOLICY

5 deactivation of audi t pol icies

CREATE STRUCTUREDPRIVILEGE

6 creation of s tructured/ana lytica l privi leges

DROP STRUCTUREDPRIVILEGE

6 destruction of s tructured/ana lytica l privi lege

ALTER STRUCTUREDPRIVILEGE

6 change of s tructured/ana lytica l privi lege

SET SYSTEM LICENSE 7 insta l la tion of a sys tem l i cense

UNSET SYSTEMLICENSE

7 deletion of l i censes

DROP TABLE 7 deletion of database tables

REPOSITORYACTIVATE

7 activation of repos i tory des ign time objects

INSERT 7use of insert/replace/upsert s tatements ontables and views

a l lows speci fi cation oftarget objects

UPDATE 7use of update/replace/upsert s tatements ontables and views


DELETE 7deletion of rows from tables/views andtruncation of tables


SELECT 7 use of select s tatements on tables and viewsa l lows speci fi cation oftarget objects

EXECUTE 7 procedure ca l l sa l lows speci fi cation oftarget objects

ALL 7 a l l actions abovetypica l ly us ed forspeci fi c us ers

Configuration Parameter

Currently the configuration parameter for auditing are stored in global.ini configuration file, in the auditingconfiguration section.

global_auditing_state ( 'true' / 'false' )

Activates / deactivates auditing globally, regardless of the enabled state of the audit policies. The default is


80 2012-12-19

false, meaning: no auditing will occur.

default_audit_trail_type ( 'SYSLOGPROTOCOL' / 'CSVTEXTFILE' )

Specifies how to store the auditing results. SYSLOGPROTOCOL is the default. CSVTEXTFILE should be used only for testing purposes.

default_audit_trail_path

Specifies the audit file storage location for the CSVTEXTFILE audit trail type.

As for all configuration parameters, these parameters can be selected in view M_INIFILE_CONTENTS,assuming that the current user has the required privileges.

Note:These parameters can only be seen if they have been explicitly set.

System and Monitoring Views

AUDIT_POLICY: shows all audit policies and their states M_INIFILE_CONTENTS: shows the configuration parameter concerning auditing

Only database users with system privilege CATALOG READ, DATA ADMIN or INIFILE ADMIN can viewinformation in the M_INIFILE_CONTENTS view. For other database users this view will be empty.

Example

Your create a new audit policy named priv_audit that will audit successful granting and revoking of privilegesand roles. The audit policy has the medium audit level CRITICAL. This policy has to be enabled explicity (see ALTER AUDIT POLICY) to make the auditing of the audit policyoccur.


You create a new audit policy named object_audit that will audit the inserts into the existing tableMY_SCHEMA.MY_TABLE. This policy has to be enabled explicity (see ALTER AUDIT POLICY) to make theauditing of the audit policy occur. This policy is restricted to user FRED and uses the audit level INFO.

CREATE USER FRED PASSWORD Initial_1; CREATE SCHEMA MY_SCHEMA OWNED BY system; CREATE TABLE MY_SCHEMA.MY_TABLE (first_col int); GRANT INSERT ON MY_SCHEMA.MY_TABLE to FRED; CREATE AUDIT POLICY OBJECT_AUDIT AUDITING SUCCESSFUL INSERT ON MY_SCHEMA.MY_TABLE FOR FRED LEVEL INFO

CREATE FULLTEXT INDEX

Syntax

CREATE FULLTEXT INDEX <index_name> ON <tableref> '(' <column_name> ')' [<fulltext_parameter_list>]

Syntax

<index_name> ::= <string_literal>

Specifies the name of the fulltext index.

<fulltext_parameter_list> ::= <fulltext_parameter> [, ...]

<fulltext_parameter> ::= LANGUAGE COLUMN <column_name> | LANGUAGE DETECTION '(' <string_literal_list> ')' | MIME TYPE COLUMN <column_name> | <change_tracking_elem> | FUZZY SEARCH INDEX <on_off> | PHRASE INDEX RATIO <on_off> | CONFIGURATION <string_literal> | SEARCH ONLY <on_off> | FAST PREPROCESS <on_off>


81 2012-12-19

<on_off> ::= ON | OFF

LANGUAGE COLUMNColumn where the language of a document is specified.

LANGUAGE DETECTIONSet of languages to be considered for language detection.

MIME TYPE COLUMNColumn where the mime-type of a document is specified.

FUZZY SEARCH INDEXSpecifies whether the fuzzy search index should be used.

PHRASE INDEX RATIOSpecifies the percentage of the phrase index. Value must be between 0.0 and 1.0

CONFIGURATIONPath to a custom configuration file for text analysis.

SEARCH ONLYIf set to ON the original document content is not stored.

FAST PREPROCESSIf set to ON, fast preprocessing is used, i.e. linguistic searches are not possible.

<change_tracking_elem> ::= SYNC[HRONOUS] | ASYNC[HRONOUS] [FLUSH [QUEUE] <flush_queue_elem>]

SYNCIf specified a synchronous fulltext index is created.

ASYNCIf specified an asynchronous fulltext index is created.

<flush_queue_elem> ::= EVERY <integer_literal> MINUTES | AFTER <integer_literal> DOCUMENTS | EVERY <integer_literal> MINUTES OR AFTER <integer_literal> DOCUMENTS

Specifies when to update the fulltext index if an asynchronous index is used.

Description

The CREATE FULLTEXT INDEX statement creates an explicit fulltext index on the given table column.

Example

You create table A.

CREATE COLUMN TABLE A (A VARCHAR(10) PRIMARY KEY, B VARCHAR(10));

You create synchronous fulltext index named 'i'. Index i is on column 'A' of table 'A', where the fuzzy searchindex is not used and the set of languages for language detection consists of 'EN','DE' and 'KR'.

CREATE FULLTEXT INDEX i ON A(A) FUZZY SEARCH INDEX OFF SYNC LANGUAGE DETECTION ('EN','DE','KR')

CREATE INDEX

Syntax

CREATE [UNIQUE] [BTREE | CPBTREE] INDEX <index_name> ON <table_name> (<column_name_order>, ...) [ASC | DESC]

Syntax Elements

<index_name> ::= [<schema_name>.]<identifier> <column_name_order> ::= <column_name> [ASC | DESC]


82 2012-12-19

UNIQUEUsed to create unique indexes. Check for duplicates will occur when an index is created and when a record isadded to the table.

BTREE | CPBTREEUsed to select the kind of index to use.

BTREE keyword has to be used in order to use B+-tree index and the CPBTREE keyword has to be used for theCPB+-tree index.B+-tree is a tree which maintains sorted data that performs efficient insertion, deletion and search ofrecords.CPB+-tree stands for Compressed Prefix B+-Tree, which is based on pkB-tree. CPB+-tree is a very small indexbecause it uses 'partial key' that is only part of full key in index nodes. CPB+-tree shows better performancethan B+-Tree for larger keys.When column data types are character string types, binary string types, decimal types, or when theconstraint is a composite key, or a non-unique constraint, the default index type is CPBTREE; otherwise,BTREE is used.If neither BTREE nor CPBTREE keyword is specified, then SAP HANA database chooses the appropriate indextype.

ASC | DESCSpecifies whether the index should be created in ascending or descending order.These keywords can be only used in the btree index, and can be specified only once for each column.

Description

The CREATE INDEX statement creates an index.

Example

Table t is created, then CPBTREE index idx is created on column b of table t with ascending order:

CREATE TABLE t (a INT, b NVARCHAR(10), c NVARCHAR(20)); CREATE INDEX idx ON t(b);

CPBTREE index idx1 is created on column a of table t with ascending order and column b with descedingorder:

CREATE CPBTREE INDEX idx1 ON t(a, b DESC);

CPBTREE index idx2 is created on column a and c of table t with desceding order:

CREATE INDEX idx2 ON t(a, c) DESC;

UNIQUE CPBTREE index idx3 is created on column b and c of table t with ascending order:

CREATE UNIQUE INDEX idx3 ON t(b, c);

UNIQUE BTREE index idx4 is created on column a of table t with ascending order:

CREATE UNIQUE INDEX idx4 ON t(a);

CREATE SCHEMA

Syntax

CREATE SCHEMA <schema_name> [OWNED BY <user_name>]

Syntax Elements

<schema_name> ::= <identifier> <user_name> ::= <simple_identifier>

OWNED BYSpecifies the name of the schema owner. If omitted, the current user will be the owner of the schema.

Description

The CREATE SCHEMA statement creates a schema in the current database.


83 2012-12-19

The CREATE SCHEMA statement creates a schema in the current database.

Example

CREATE SCHEMA my_schema OWNED BY system;

CREATE SEQUENCE

Syntax

CREATE SEQUENCE <sequence_name> [<common_sequence_parameter_list>] [RESET BY <subquery>]

Syntax Elements


<common_sequence_parameter_list> ::= <common_sequence_parameter>, ... <common_sequence_parameter> ::= <sequence_parameter_start_with> | <basic_sequence_parameter> <basic_sequence_parameter> ::= INCREMENT BY n | MAXVALUE n | NO MAXVALUE | MINVALUE n | NO MINVALUE | CYCLE | NO CYCLE <sequence_parameter_start_with> ::= START WITH n

INCREMENT BYDefines the amount the next sequence value is incremented from the last value assigned. The default is 1.Specify a negative value to generate a descending sequence. An error is returned if the INCREMENT BY valueis 0

START WITHDefines the starting sequence value. If you do not specify a value for the START WITH clause, MINVALUE isused for ascending sequences and MAXVALUE is used for descending sequences.

MAXVALUEDefines the largest value generated by the sequence and must be between 0 and 4611686018427387903.

NO MAXVALUEWhen MAXVALUE is not specified, the maximum value for an ascending sequence is 4611686018427387903and the maximum value for a descending sequences is -1.

MINVALUEThe minimum value of a sequence can be specified after MINVALUE and is between 0 and4611686018427387903.

NO MINVALUEWhen MINVALUE is not specified, the minimum value for an ascending sequence is 1 and the minimum valuefor a descending is -4611686018427387903.

CYCLEThe sequence number will be reused after it reaches its maximum or minimum value.

NO CYCLEDefault option. The sequence number will not be reused after it reaches its maximum or minimum value.

RESET BYDuring the restart of the database, database automatically executes the RESET BY statement and thesequence value is restarted with the specified value from the statement after RESET BY.If RESET BY is not specified, the sequence value is stored persistently in database. During the restart of thedatabase, the next value of the sequence is generated from the saved sequence value.

Description

The CREATE SEQUENCE statement is used to create a sequence.

A sequence is used to generate unique integers by multiple users. CURRVAL is used to get the current valueof the sequence and NEXTVAL is used to get the next value of the sequence. CURRVAL is only valid aftercalling NEXTVAL in a session.


84 2012-12-19

Example

Example 1:Sequence seq is created, then CURRVAL and NEXTVAL are used to get the values from the sequence:

CREATE SEQUENCE seq START WITH 11;

NEXTVAL returns 11:

SELECT seq.NEXTVAL FROM DUMMY;

CURRVAL returns 11:

SELECT seq.CURRVAL FROM DUMMY;

Example 2:If the sequence s is used to create a unique key on column A in the table R, then after a database is restarted,a UNIQUE key value can be created by automatically assigning the maximum value of column A to thesequence value using a RESET BY statement as follows:

CREATE TABLE r (a INT); CREATE SEQUENCE s RESET BY SELECT IFNULL(MAX(a), 0) + 1 FROM r; SELECT s.NEXTVAL FROM DUMMY;

CREATE SYNONYM

Syntax

CREATE [PUBLIC] SYNONYM <synonym_name> FOR <object_name>

Syntax Elements

<synonym_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The name of the synonym to be created, with optional schema name.

<object_name> ::= <table_name> | <view_name> | <procedure_name> | <sequence_name>

<table_name> ::= [<schema_name>.]<identifier> <view_name> ::= [<schema_name>.]<identifier> <sequence_name> ::= [<schema_name>.]<identifier> <procedure_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

Description

The CREATE SYNONYM creates an alternate name for a table, view, procedure or sequence.

You can use a synonym to re-point functions and stored procedures to differing tables, views or sequenceswithout needing to re-write the function or stored procedure.

The optional PUBLIC element allows for the creation of a public synonym. A public synonym can be accessedby any users, but the users that have proper privilege on its base object can access the base object.

Example

You create table A.

CREATE TABLE A (A INT PRIMARY KEY, B INT);

You create a synonym for table A, a_synonym.

CREATE SYNONYM a_synonym FOR A;


85 2012-12-19

CREATE TABLE

Syntax

CREATE [<table_type>] TABLE <table_name> <table_contents_source> [<logging_option>] [<auto_merge_option>] [<unload_priority_clause>] [<schema_flexibility_option>] [<partition_clause>] [<location_clause>]

Syntax elements:<table_type> <table_contents_source><logging_option><auto_merge_option><unload_priority_clause><schema_flexibility_option><partition_clause><location_clause>

Syntax Elements

<table_type>

<table_type> ::= COLUMN | ROW | HISTORY COLUMN | GLOBAL TEMPORARY | LOCAL TEMPORARY

Defines the type of table storage organization. The default value is ROW.

ROW, COLUMN

If the majority of table access will be through a large number of tuples, with only a few selected attributes,COLUMN-based storage should be used. If the majority of table access involves selecting a few records, withall attributes selected, ROW-based storage is preferable. The SAP HANA database uses a combination oftable types to enable storage and interpretation in both ROW and COLUMN forms.

HISTORY COLUMN

Creates a table with a particular transaction session type called 'HISTORY'. Tables with session type HISTORYsupport time travel queries. Time travel queries are queries against historical states of the database.

For more information on utilizing history time travel please see: HISTORY time travel

GLOBAL TEMPORARY

Table definition is globally available while data is visible only to the current session. The table is truncated atthe end of the session.

Metadata in a global temporary table is persistent meaning the metadata exists until the table is droppedand the metadata is shared across sessions. Data in a global temporary table is session-specific meaning onlythe owner session of the global temporary table is allowed to insert/read/truncate the data, exists for theduration of the session and data from the global temporary table is automatically dropped when the sessionis terminated. Global temporary table can be dropped only when the table does not have any record in it.

Supported operations on Global Temporary Table:

1. Create without a primary key2. Rename table3. Rename column4. Truncate5. Drop6. Create or Drop view on top of global temporary table7. Create synonym8. Select9. Select into or Insert

10. Delete11. Update12. Upsert or Replace


86 2012-12-19

LOCAL TEMPORARY

The table definition and data is visible only to the current session. The table is truncated at the end of thesession.

Metadata exists for the duration of the session and is session-specific meaning only the owner session of thelocal temporary table is allowed to see. Data in a local temporary table is session-specific meaning only theowner session of the local temporary table is allowed to insert/read/truncate the data, exists for the durationof the session and data from the local temporary table is automatically dropped when the session isterminated.

Supported operations on Local Temporary Table:

1. Create without a primary key2. Truncate3. Drop4. Select5. Select into or Insert6. Delete7. Update8. Upsert or Replace


For descriptions on schema name and identifier, please refer to Identifiers.

<table_contents_source>

<table_contents_source> ::= (<table_element>, ...) | <like_table_clause> | [(<column_name>, ...)] <as_table_subquery>

You can create a table in the following ways:

Describing the table elements, see below.Basing the new table like an existing table, see CREATE TABLE LIKE.From the result of a subquery, see CREATE TABLE AS <subquery>.

<table_element> ::= <column_definition> [<column_constraint>] | <table_constraint>

Defines a table column with associated column or table constraint. For column constraint please see column constraint.For table constraint please see table constraint.

<column_definition> ::= <column_name> {<data_type> | <lob_data_type>} [<column_store_data_type>] [<ddic_data_type>] [DEFAULT <default_value>] [GENERATED ALWAYS AS <expression>] [<schema_flexibility>] [<fuzzy_search_index>] [<fuzzy_search_mode>]




<data_type> ::= DATE | TIME | SECONDDATE | TIMESTAMP | TINYINT | SMALLINT | INTEGER | BIGINT | SMALLDECIMAL | DECIMAL | REAL | DOUBLE | VARCHAR | NVARCHAR | ALPHANUM | SHORTTEXT | VARBINARY | TEXT

<lob_data_type> ::= <lob_type_name> [MEMORY THRESHOLD <memory_threshold_value>] <lob_type_name> ::= BLOB | CLOB | NCLOB


<ddic_data_type> ::= DDIC_ACCP | DDIC_ALNM | DDIC_CHAR | DDIC_CDAY | DDIC_CLNT | DD


87 2012-12-19

IC_CUKY | DDIC_CURR | DDIC_D16D | DDIC_D34D | DDIC_D16R | DDIC_D34R | DDIC_D16S | DDIC_D34S | DDIC_DATS | DDIC_DAY | DDIC_DEC | DDIC_FLTP | DDIC_GUID | DDIC_INT1 | DDIC_INT2 | DDIC_INT4 | DDIC_INT8 | DDIC_LANG | DDIC_LCHR | DDIC_MIN | DDIC_MON | DDIC_LRAW | DDIC_NUMC | DDIC_PREC | DDIC_QUAN | DDIC_RAW | DDIC_RSTR | DDIC_SEC | DDIC_SRST | DDIC_SSTR | DDIC_STRG | DDIC_STXT | DDIC_TIMS | DDIC_UNIT | DDIC_UTCM | DDIC_UTCL | DDIC_UTCS | DDIC_TEXT | DDIC_VARC | DDIC_WEEK

The available data types. Please see Data Types.

<memory_threshold_value> ::= <unsigned_integer> | NULL

Controls if LOB data should be stored in memory or not according to the following conditions:

If <memory_threshold_value> is not provided, lob is stored in memory by default.If <memory_threshold_value> is provided and its LOB size is bigger than memory threshold value, LOBdata is stored on disk.If <memory_threshold_value> is provided and its LOB size is equal or less than memory thresholdvalue, LOB data is stored in memory.If <memory_threshold_value> is NULL, all LOB data is stored in memory.If <memory_threshold_value> is 0, all LOB data is stored in disk.

DEFAULT <default_value> ::= NULL | <string_literal> | <signed_numeric_literal> | <unsigned_numeric_literal>

Specifies a value to be assigned to the column if an INSERT statement does not provide a value for thecolumn.

GENERATED ALWAYS AS <expression>

Specifies the expression to generate the column value in runtime.

LOB DATA TYPE in column definitionAvaliable lob data types are BLOB, CLOB and NCLOB.If MEMORY THRESHOLD option is not given, lob is stored in memory as default.If MEMORY THRESHOLD is given and its LOB size is bigger than memory threshold value, LOB data is storedin disk. If MEMORY THRESHOLD is given and its LOB size is equal or less than memory threshold value, LOBdata is stored in memory. If memory threshold is NULL, all LOB data is stored in memory. If memorythreshold is 0, all LOB data is stored in disk.

<schema_flexibility> ::= {ENABLE | DISABLE} SCHEMA FLEXIBILITY

Specifies the column is dynamic:

ENABLE produces a dynamic column.DISABLE (default value) produces a static column.

<fuzzy_search_index> ::= FUZZY SEARCH INDEX {ON | OFF}

Turns a fuzzy search index on or off. OFF is the default.

<fuzzy_search_mode> ::= FUZZY SEARCH MODE [<string_literal> | NULL]

Sets the fuzzy search mode with the value of <string_literal>. If NULL is specified, the fuzzy search mode isreset.

<column_constraint> ::= NULL | NOT NULL | <unique_specification> | <references_specification>

The column constraint rules.For NULL, NOT NULL and <unique_specification>, see below. For <references_specification>, see References Specification

NULL

If NULL is specified it is not considered a constraint, it represents that a column that may contain a null value.


88 2012-12-19

The default is NULL.

NOT NULL

The NOT NULL constraint prohibits a column value from being NULL.

<unique_specification> ::= UNIQUE [<unique_tree_type_index>] | PRIMARY KEY [<unique_tree_type_index>]

Specifies unique constraints. If the index type is omitted, the SAP HANA database chooses the appropriateindex by considering the column data type. If the index type is not specified, the SAP HANA database willautomatically select an index type as follows:

Index type Criteria

CPBTREE

- character s tring types . - binary s tring types .- decimal types .- when the constra int i s a compos i te key.- when the constra int i s a non-unique constra int

BTREE Al l other cases than speci fied for CPBTREE

UNIQUE

Specifies a column as a unique key.A composite unique key enables the specification of multiple columns as a unique key. With a uniqueconstraint, multiple rows cannot have the same value in the same column.

PRIMARY KEY

A primary key constraint is a combination of a NOT NULL constraint and a UNIQUE constraint. It prohibitsmultiple rows from having the same value in the same column.

<unique_tree_type_index> ::= BTREE | CPBTREE;

Specifies the index type.

BTREE

BTREE specifies a B+-tree index. B+-tree is a tree that maintains sorted data which performs efficientinsertion, deletion and search of records.

CPBTREE

CPBTREE specifies a CPB+-tree index. CPB+-tree stands for Compressed Prefix B+-Tree, which is based on pkB-tree. CPB+-tree is a very small index because it uses 'partial key' that is only part of full key in index nodes.CPB+-tree shows better performance than B+-Tree for larger keys.

<table_constraint> ::= <unique_constraint_definition> | <referential_constraint_definition>

The table constraint can be either a unique constraint or a referential constraint.

<unique_constraint_definition> ::= <unique_specification> (<unique_column_name_list>)

The unique specification see Unique specifcation

<unique_column_name_list> ::= <unique_column_name>[{, <unique_column_name>}...]

Speicfies the unique column name list which can have one or more column names.

<unique_column_name> ::= <identifier>

A column name identifier.

<referential_constraint_definition> ::= FOREIGN KEY (<referencing_column_name_list>) <references_specification>



89 2012-12-19


<referencing_column_name_list> ::= <referencing_column_name>[{, <referencing_column_name>}...]

Specifies the referencing column name list which can have one or more column names.

<referencing_column_name> ::= <identifier>

The identifier of a referencing column.

<references_specification> ::= REFERENCES <referenced_table> [(<referenced_column_name_list>)] [<referential_triggered_action>]

Specifies the referenced table, with optional column name list and trigger action. If<referenced_column_name_list> is specified, there will be one-to-one correspondence between<column_name> of <column_definition> (see column defintion) and <referenced_column_name>. If it is notspecified, there will be one-to-one correspondence between <column_name> of <column_definition> and thecolumn name of the referenced table's primary key.

<referenced_column_name_list> ::= <referenced_column_name>[{, <referenced_column_name>}...]

Specifies the referenced column name list, which can have one or more column names.

<referenced_table> ::= <identifier>

The indentifier of a table to be referenced.

<referenced_column_name> ::= <identifier>

The identifier of the column name to be referenced.

<referential_triggered_action> ::= <update_rule> [<delete_rule>] | <delete_rule> [<update_rule>]

Specifies an update rule with optional delete rule or a delete rule with optional update rule. The order inwhich they are provided provides an order of precedence for execution.

<update_rule> ::= ON UPDATE <referential_action> <referential_action> ::= CASCADE | RESTRICT | SET DEFAULT | SET NULL

The following UPDATE referential actions are possible:

ActionName

Update Action

RESTRICTAny updates to a referenced table are prohibi ted i f there are any matched records in thereferencing table. This i s the defaul t action.

CASCADEIf a record i s updated in the referenced table, the corres ponding records in the referencing tableare a lso updated with the same va lues .

SET NULLIf a record i s updated in the referenced table, the corres ponding records in the referencing tableare a lso updated with nul l va lues .

SETDEFAULT

If a record i s updated in the referenced table, the corres ponding records in the referencing tableare a lso updated with thei r defaul t va lues

<delete_rule> ::= ON DELETE <referential_action>

The following DELETE referential actions are possible:


90 2012-12-19

ActionName

Delete Action

RESTRICTAny deletions to a referenced table are prohibi ted i f there are any matched records in thereferencing table. This i s the defaul t action.

CASCADEIf a record in the referenced table i s deleted, the corres ponding records in the referencing tableare a lso deleted.

SET NULLIf a record in the referenced table i s deleted, the corres ponding records in the referencing tableare s et to nul l .

SETDEFAULT

If a record in the referenced table i s deleted, the corres ponding records in the referencing tableare s et to thei r defaul t va lues .

<like_table_clause> ::= LIKE <like_table_name> [WITH [NO] DATA] [<like_without_option>]

Creates a table that has the same definition as like_table_name, unless corresponding property or WITHOUToption is specified. All the column definitions with constraints, default values and other properties such asgenerated column, schema flexibility, etc are copied from the table <like_table_name>. All table constraints,indexes, fulltext indexes and table location are also copied.

<like_table_name> ::= <table_name> <table_name> ::= <identifier>

The identifier of the table that will be duplicated.

WITH [NO] DATA

The WITH DATA option specifies that data is copied from the <like_table_name> table. The default value ofWITH NO DATA does not copy the table data.

<like_without_option> ::= WITHOUT AUTO MERGE | WITHOUT HISTORY | WITHOUT NO LOGGING | WITHOUT PARTITION | WITHOUT SCHEMA FLEXIBILITY | WITHOUT UNLOAD | WITHOUT PRELOAD | WITHOUT UNLOAD PRIORITY | WITHOUT INDEX | WITHOUT FUZZY SEARCH INDEX | WITHOUT FUZZY SEARCH MODE

Specifies which properties will not be copied from the <like_table_name> table. For example when WITHOUTPARTITION is specified, the table will not be partitioned.

<as_table_subquery> ::= AS (<subquery>) [WITH [NO] DATA]

Creates a table and fills it with the data computed by the <subquery> (See sql_subquery). Only NOT NULLconstraints are copied by this clause. If column_names are specified, specified column_names override thecolumn names from <subquery>.

WITH [NO] DATA

The WITH DATA option specifies that data is copied from <subquery>. The default is WITH DATA. WITH NODATA option does not copy the data from <subquery>.

<logging_option>

<logging_option> ::= LOGGING | NO LOGGING [RETENTION <retention_period>]

Specifies logging options for the created table.

LOGGING

Specifies that table logging is activated. This is the default logging option.

NO LOGGING

Specifies that logging is deactivated. A NO LOGGING table means that the definition of the table is persistent


91 2012-12-19

and globally available, data is temporary and global.

RETENTION <retention_period> <retention_period> ::= <unsigned_integer>

Specifies the retention time in seconds of the column table created by NO LOGGING. After the specifiedretention period has elapsed, the table will be dropped if used physical memory of the host is above 80%.

<auto_merge_option>

<auto_merge_option> ::= [NO] AUTO MERGE

AUTO MERGE (default value) specifies that automatic delta merge is triggered. NO AUTO MERGE disablesautomatic delta merging.

<unload_priority_clause>

<unload_priority_clause> ::= UNLOAD PRIORITY <unload_priority>

UNLOAD PRIORITY specifies that priority of table to be unloaded from memory.

<unload_priority> ::= <digit>

<unload_priority> can be 0 ~ 9, where 0 means not-unloadable and 9 means earliest unload.

<schema_flexibility_option>

<schema_flexibility_option> ::= WITH SCHEMA FLEXIBILITY

Specifies that the table schema is flexible.

<partition_clause>

<partition_clause> ::= PARTITION BY <hash_partition> [, <range_partition> | , <hash_partition>] | PARTITION BY <range_partition> | PARTITION BY <roundrobin_partition> [,<range_partition>]

The PARTITION BY option partitions a table using the selected rules. For more information about partitioning please see the "Table Partitioning in the SAP HANA database"section in the "SAP HANA Administration guide" available from the SAP HANA Appliance page.

<hash_partition> ::= HASH (<partition_expression> [{<partition_expression>,}...]) PARTITIONS {<num_partitions> | GET_NUM_SERVERS()}

Partitions the created table using a hash partitioning scheme.

<range_partition> ::= RANGE (<partition_expression>) (<range_spec>, ...)

Partitions the created table using a range partitioning scheme.

<roundrobin_partition> ::= ROUNDROBIN PARTITIONS {<num_partitions> | GET_NUM_SERVERS()} [, <range_partition>]

Partitions the created table using a round robin partitioning scheme.

GET_NUM_SERVERS()

Returns the number of servers available in the partitioning cluster.

<range_spec> ::= {<from_to_spec> | <single_spec>} [{{<from_to_spec> | <single_spec>},} ...] [, PARTITION OTHERS]

The range specifier for a partition.

<from_to_spec> ::= PARTITION <lower_value> <= VALUES < <upper_value>

Specifies a partition using lower and upper values of a <partition_expression>.


92 2012-12-19


<single_spec> ::= PARTITION VALUE = <target_value>

Specifies a partition using a single value of a <partition_expression>.

PARTITION OTHERS

Specifies that all other values that are not covered by the partition specification will be gathered into onepartition.

<partition_expression> ::= <column_name> | YEAR(<column_name>) | MONTH(<column_name>)

The specifier used to segregate data into partitions.

<lower_value> ::= <string_literal> | <numeric_literal>

The lower value of a partition specifier.

<upper_value> ::= <string_literal> | <numeric_literal>

The upper value of a partition specifier.

<target_value> ::= <string_literal> | <numeric_literal>

The target value of a single partition specifier.

<num_partitions> ::= <unsigned_integer>

The number of partitions to be created for the table.

<location_clause>

<location_clause> ::= AT [LOCATION] {'<host>:<port>' | ('<host>:<port>', ...)} <host> ::= <string_literal> <port> ::= <unsigned_integer>

It is possible to determine the index servers on which the partitions are created. If you specify the LOCATION,the partitions will be created on these instances using round robin scheme. Duplicates in the list will beremoved. If you specify exactly the same number of instances as partitions in the partition specification, theneach partition will be assigned to the respective instance in the list. All index servers in the list have to belongto the same instance.

If no locations are specified, the partitions will be created randomly. If the number of partitions matches thenumber of servers - for example by using GET_NUM_SERVERS() - it is ensured that multiple CREATE TABLEcalls will be distributed across all servers evenly. In case of a multi-level partitioning, this applies for thenumber of partitions of the first level.

This mechanism is useful if several tables are to be created which have a semantic relationship to each other.A table can be created in the specified location with <host>:<port>. <location_clause> can be specified whencreating partitioned tables that are distributed on multiple instances. When <location_clause> is providedwithout <partition_clause>, the table is created on the first location specified. If location information is notprovided, the table will be automatically assigned to one node. This option can be used for both row storeand column store tables in a distributed environment.

Description

The CREATE TABLE statement creates a table. Tables are created without data except when<as_table_subquery> or <like_table_clause> is used along with the WITH DATA option.

Examples

You create Table A that has INTEGER-type column A and B. Column A has a primary key constraint.


You create a partitioned table P1 that has DATE-type column U. Column U has a primary key constraint and isused as RANGE-partitioning column.

CREATE COLUMN TABLE P1 (U DATE PRIMARY KEY) PARTITION BY RANGE (U) (PARTITION '2010


93 2012-12-19

-02-03' <= VALUES < '2011-01-01', PARTITION VALUE = '2011-05-01');

You create a partitioned table P2 that has INTEGER-type column I, J and K. Column I and J make the primarykey constraint and are used as HASH-partitioning column. Column K is used as a sub HASH-partitioningcolumn.

CREATE COLUMN TABLE P2 (I INT, J INT, K INT, PRIMARY KEY(I, J)) PARTITION BY HASH (I, J) PARTITIONS 2, HASH (K) PARTITIONS 2;

You create table C1 that has the same definition as table A. Table C1 also has the same records as table A.

CREATE COLUMN TABLE C1 LIKE A WITH DATA;

You create table H1 that has the same definition as table H. Table H1 also has the same records as table H.You create a column table H2 from the definition of table H without history property and without anyrecords.

CREATE HISTORY COLUMN TABLE H (A INT); CREATE TABLE H1 LIKE H WITH DATA; CREATE TABLE H2 LIKE H WITH NO DATA WITHOUT HISTORY;

You create table C2 that has the same column data type and NOT NULL constraint as table A without anydata.

CREATE TABLE C2 AS (SELECT * FROM A) WITH NO DATA;

You create table F that has a foreign key that references column A of table R with update cascade option.

CREATE TABLE R (A INT PRIMARY KEY, B NVARCHAR(10)); CREATE TABLE F (FK INT, B NVARCHAR(10), FOREIGN KEY(FK) REFERENCES R ON UPDATE CASCADE);

You create a flexible table T.

CREATE COLUMN TABLE T(A INT) WITH SCHEMA FLEXIBILITY;

You create table T2 with fuzzy search index and a fuzzy search mode.

CREATE COLUMN TABLE T2 (KEY INT, COL1 VARCHAR(10) FUZZY SEARCH INDEX ON, COL2 NVARCHAR(10) FUZZY SEARCH MODE 'POSTCODE');

HISTORY time travel

You can utlitize the HISTORY COLUMN table type time travel feature in the following ways:

Session level time travel

A database session can be set back to a certain point-in-time. The input parameter of the statement is eithera utc-timestamp or a commitid.

SET HISTORY SESSION TO UTCTIMESTAMP = <timestamp> <timestamp> ::= <string_literal>

or

SET HISTORY SESSION TO COMMIT ID = <commit_id> <commit_id> ::= <unsigned_integer>

The timestamp has to be in format 'YYYY-MM-DD HH:MM:SS[.FF7]'. The COMMIT ID variant of thestatement takes a commitid as a parameter. The value of the commitid parameter must occur inCOMMIT_ID column of the system table SYS.TRANSACTION_HISTORY, otherwise an exception will bethrown. The COMMIT ID is useful when using user defined snapshots. A user defined snapshot can be takenby simply storing the commitid which is assigned to a transaction during the commit phase. The commitid canbe retrieved by executing the following query directly after a transaction commit:

SELECT LAST_COMMIT_ID FROM M_TRANSACTIONS; WHERE CONNECTION_ID = CURRENT_CONNECTION;

The TIMESTAMP-variant of the statement takes a timestamp as parameter. Internally, the timestamp isused to look up a (commit_time,commit_id)-pair inside the system-table SYS.TRANSACTION_HISTORY where


94 2012-12-19

the commit_time is close to the given timestamp (to be more precisely: choose pair where maximalCOMMIT_TIME is smaller or equal to the given timestamp; if no such pair is found an exception will beraised). The session then will be restored with the determined commit-id as in the COMMIT ID variant.To terminate a restored session to switch back to the current session, an explicit COMMIT or ROLLBACK hasto be executed on the DB connection.

Statement level time travel

<subquery> AS OF UTCTIMESTAMP <timestamp> <subquery> AS OF COMMIT ID <commit_id>

In order to be able to relate the commitid with the commit time, a system table SYS.TRANSACTION_HISTORYis maintained which stores additional information for each transaction which commits data for history table.For detailed information on setting session-level timetravel, please refer to SET HISTORY SESSION, and fordetails on <subquery>, please refer to Subquery.

Note:

Autocommit has to be turned off when a session should be restored (otherwise an exceptionwill be thrown with an appropriate error message)Non-history tables in restored sessions always show their current snapshotOnly data query statement (select) is allowed inside restored sessions.A history table must have a primary keyThe session type can be checked from the column, SESSION_TYPE of the system tableSYS.TABLES.

CREATE TRIGGER

Syntax

CREATE TRIGGER <trigger_name> <trigger_action_time> <trigger_event> ON <subject_table_name> [REFERENCING <transition_list>] [<for_each_row>] BEGIN [<trigger_decl_list>] [<proc_handler_list>] <trigger_stmt_list> END

Syntax Elements

<trigger_name> ::= <identifier>

The name of trigger you create.

<subject_table_name> ::= <identifier>

The name of table your trigger is defined on. For descriptions on identifier, see Identifiers.

<trigger_action_time> ::= BEFORE | AFTER

Specify when trigger acction occur.

BEFORETrigger is executed before the subject table operation.

AFTERTrigger is executed after the subject table operation.

<trigger_event> ::= INSERT | DELETE | UPDATE

Specify which data modification command activate trigger action.

<transition_list> ::= <transition> | <transition_list> , <transition> <transition> ::= <trigger_transition_old_or_new> <trigger_transition_var_or_table> <trans_var_name> | <trigger_transition_old_or_new> <trigger_transition_var_or_table> AS <trans_var_name>

When trigger transition variable is declared, trigger can access records that are being changed by trigger


95 2012-12-19

triggering DML.While row level trigger execution, <trans_var_name>.<column_name> represents corresponding columnrecord being changed in trigger.Here <column_name> is subject table's column name. See example of transition variable

<trigger_transition_old_or_new> ::= OLD | NEW <trigger_transition_var_or_table> ::= ROW <trans_var_name> ::= <identifier>

OLD You can access old record of DML in trigger, which is to-be-replaced-by-update old record or deleted-old-record. UPDATE trigger and DELETE trigger can have OLD ROW transition variable.

NEWYou can access new record of DML in trigger, which is to-be-inserted new record or to-be-updated newrecord.UPDATE trigger and INSERT trigger can have NEW ROW transition variable.

Only transition variable is supported.Transition table is not supported feature. If you put 'TABLE' as <trigger_transition_var_or_table>, you will see not-supported-feature error.

<for_each_row> ::= FOR EACH ROW

Defines that the trigger will be called in a row-wise fashion. Statement level trigger execution is the default mode when FOR EACH ROW is not defined.Currently statement-wise trigger is not supported.

<trigger_decl_list> ::= DECLARE <trigger_decl> | <trigger_decl_list> DECLARE <trigger_decl> <trigger_decl> ::= <trigger_var_decl> | <trigger_condition_decl> <trigger_var_decl> ::= <var_name> CONSTANT <data_type> [<not_null>] [<trigger_default_assign>] ; | <var_name> <data_type> [<not_null>] [<trigger_default_assign>] ; <var_name> ::= <identifier> <data_type> ::= DATE | TIME | SECONDDATE | TIMESTAMP | TINYINT | SMALLINT | INTEGER | BIGINT | SMALLDECIMAL | DECIMAL | REAL | DOUBLE | VARCHAR | NVARCHAR | ALPHANUM | SHORTTEXT | VARBINARY | BLOB | CLOB | NCLOB | TEXT <not_null> ::= NOT NULL <trigger_default_assign> ::= DEFAULT <expression> | := <expression> <trigger_condition_decl> ::= <condition_name> CONDITION ; | <condition_name> CONDITION FOR <sql_error_code> ; <condition_name> ::= <identifier> <sql_error_code> ::= SQL_ERROR_CODE <int_const>

trigger_decl_listYou can declare trigger variable or condition.Declared variable can be used scalar value assignment or referenced in trigger SQL statement. Declared condition name can be referenced when exception handler.

CONSTANT When CONSTANT kewword was given, you cannot change the variable while trigger execution.

<proc_handler_list> ::= <proc_handler> | <proc_handler_list> <proc_handler> <proc_handler> ::= DECLARE EXIT HANDLER FOR <proc_condition_value_list> <trigger_s


96 2012-12-19

tmt> <proc_condition_value_list> ::= <proc_condition_value> | <proc_condition_value_list> , <proc_condition_value> <proc_condition_value> ::= SQLEXCEPTION | SQLWARNING | <sql_error_code> | <condition_name>

Exception handler can be declared for catching with existing sql exception, specific error code number or condition name declared on condition variable.

<trigger_stmt_list> ::= <trigger_stmt> | <trigger_stmt_list> <trigger_stmt> <trigger_stmt> ::= <proc_block> | <proc_assign> | <proc_if> | <proc_loop> | <proc_while> | <proc_for> | <proc_foreach> | <proc_signal> | <proc_resignal> | <trigger_sql>

Trigger body syntax is a part of procedure body syntax. See detail of create procedure definition in SAP HANA database SQLScript guide.The syntax of trigger body complies with that of procedure, which are nested block(proc_block),scalar variable assign(proc_assign), if block(proc_if), loop block(proc_loop), for block(proc_for), for each block(proc_foreach), exception signal(proc_signal), exception resignal(proc_resignal), and sql statement(proc_sql).

<proc_block> ::= BEGIN [<trigger_decl_list>] [<proc_handler_list>] <trigger_stmt_list> END ;

You can add another 'BEGIN ... END;' block in a nested way.

<proc_assign> ::= <var_name> := <expression> ;

var_name is variable name, which should be declared previously.

<proc_if> ::= IF <condition> THEN <trigger_stmt_list> [<proc_elsif_list>] [<proc_else>] END IF ;

<proc_elsif_list> ::= ELSEIF <condition> THEN <trigger_stmt_list>

<proc_else> ::= ELSE <trigger_stmt_list>

For descriptions on condition, refer to <condition> in SELECT.Using IF ... THEN ... ELSEIF ... END IF, you can control execution flow with condition.

<proc_loop> ::= LOOP <trigger_stmt_list> END LOOP ;

<proc_while> ::= WHILE <condition> DO <trigger_stmt_list> END WHILE ;

<proc_for> ::= FOR <column_name> IN [<reverse>] <expression> <DDOT_OP> <expression> DO <trigger_stmt_list> END FOR ;

<column_name> ::= <identifier> <reverse> ::= REVERSE <DDOT_OP> ::= ..

<proc_foreach> ::= FOR <column_name> AS <column_name> [<open_param_list>] DO <trigger_stmt_list> END FOR ;


97 2012-12-19

<open_param_list> ::= ( <expr_list> )

<expr_list> ::= <expression> | <expr_list> , <expression>

<proc_signal> ::= SIGNAL <signal_value> [<set_signal_info>] ;

<proc_resignal> ::= RESIGNAL [<signal_value>] [<set_signal_info>] ;

<signal_value> ::= <signal_name> | <sql_error_code>

<signal_name> ::= <identifier>

<set_signal_info> ::= SET MESSAGE_TEXT = '<message_string>'

<message_string> ::= <identifier>

SET MESSAGE_TEXTIf you set your own message with SET MESSAGE_TEXT,the message is delivered to users when specified error is thrown from trigger execution.

SIGNAL statement provide raise an exception explicitly. User-defined range (10000 ~ 19999) will be allowed for the error code to issue. RESIGNAL statement provide raise an exception on the action statement in exception handler. If error code is not specified, RESIGNAL will throw up the catched exception.

<trigger_sql> ::= <select_into_stmt> | <insert_stmt> | <delete_stmt> | <update_stmt> | <replace_stmt> | <upsert_stmt>

For descriptions on insert_stmt, refer to INSERT.For descriptions on delete_stmt, refer to DELETE.For descriptions on update_stmt, refer to UPDATE.For descriptions on replace_stmt and upsert_stmt, refer to REPLACE | UPSERT.

<select_into_stmt> ::= SELECT <select_list> INTO <var_name_list> <from_clause > [<where_clause>] [<group_by_clause>] [<having_clause>] [{<set_operator> <subquery>, ... }] [<order_by_clause>] [<limit>]

<var_name_list> ::= <var_name> | <var_name_list> , <var_name>

<var_name> ::= <identifier>

For descriptions on select_list, from_clause, where_clause, group_by_clause, having_clause, set_operator, subquery, order_by_clause, limit, refer to SELECT.

var_name is scalar variable name which should be declared before.You can assign selected item value to the scalar variable.

Description

CREATE TRIGGER statement creates a trigger.A trigger is special kind of stored procedure that automatically executes when an event occurs on a certaintable.CREATE TRIGGER command defines a set of statements that is executed when a givenoperation(INSERT/UPDATE/DELETE)takes place on a given object(subject table)

Only database users having the TRIGGER privilege for the given <subject_table_name> are allowed to createa trigger for that table. Current trigger limitation is described below:

INSTEAD_OF trigger is not supported.Access of subject table which is table that trigger is defined on, is not allowed in trigger body,


98 2012-12-19

which means any insert/update/delete/replace/select for table which trigger is on is not allowed. Only row level trigger is supported. Statement level trigger is not supported. Row level trigger means trigger action will be executed for every row changes. Statement level trigger means trigger action will be executed once for a statement execution.Syntax 'FOR EACH ROW' means row-wise trigger execution and it is default mode. Even when 'FOREACH ROW' is not defined, it is still row level trigger.Transition table (OLD/NEW TABLE) is not supported. When trigger SQL statement want to refer to data that are being changed by trigger-triggering-eventlike insert/update/delete, transition variable/table will be a way SQL statement in trigger body access old and new data. Transition variable is used by row level trigger while transition table is used by statement level trigger.Trigger execution on partitioned table to multiple host or table in slave node is not supported.Not supported trigger action feature(while procedure does) : resultset assignment(select resultset assignment to tabletype), exit/continue command(execution flow control), cursor open/fetch/close(get each record data of search result by cursor and access record in loop), procedure call(call another proecedure), dynamic sql execution(build SQL statements dynamically at runtime of SQLScript), return(end SQL statement execution)


TRIGGERS is system view for trigger.

System view TRIGGERS shows: SCHEMA_NAME, TRIGGER_NAME, TRIGGER_OID, OWNER_NAME, OWNER_OID,SUBJECT_TABLE_SCHEMA, SUBJECT_TABLE_NAME, TRIGGER_ACTION_TIME, TRIGGER_EVENT, TRIGGERED_ACTION_LEVEL,DEFINITION

Example

You first need table which trigger is defined on.

CREATE TABLE TARGET ( A INT);

You also need tables which trigger accesses and modifies.

CREATE TABLE SAMPLE ( A INT);

Following is example of create trigger.

CREATE TRIGGER TEST_TRIGGER AFTER INSERT ON TARGET FOR EACH ROW BEGIN DECLARE SAMPLE_COUNT INT; SELECT COUNT(*) INTO SAMPLE_COUNT FROM SAMPLE; IF :SAMPLE_COUNT = 0 THEN INSERT INTO SAMPLE VALUES(5); ELSEIF :SAMPLE_COUNT = 1 THEN INSERT INTO SAMPLE VALUES(6); END IF; END;

Trigger TEST_TRIGGER will be executed after any record insert execution for TARGET table.Since SAPMLE table record count is zero in the first insert try, the trigger TEST_TRIGGER will insert 5 toSAMPLE table.In the second insertion to TARGET table, the trigger insert 6 to SAMPLE table because its count is two.

INSERT INTO TARGET VALUES (1); SELECT * FROM SAMPLE; 5 INSERT INTO TARGET VALUES (2); SELECT * FROM SAMPLE; 5 6

Followings are more example of create trigger. FOR/WHILE Example


99 2012-12-19

CREATE TABLE TARGET ( A INT); CREATE TABLE SAMPLE ( A INT); CREATE TRIGGER TEST_TRIGGER_WHILE_UPDATE AFTER UPDATE ON TARGET BEGIN DECLARE found INT := 1; DECLARE val INT := 1; WHILE :found <> 0 DO SELECT count(*) INTO found FROM sample WHERE a = :val; IF :found = 0 THEN INSERT INTO sample VALUES(:val); END IF; val := :val + 1; END WHILE; END;

CREATE TABLE TARGET ( A INT); CREATE TABLE control_tab(id INT PRIMARY KEY, name VARCHAR(30), payment INT); CREATE TABLE message_box(message VARCHAR(200), log_time TIMESTAMP);

CREATE TRIGGER TEST_TRIGGER_FOR_INSERT AFTER INSERT ON TARGET BEGIN DECLARE v_id INT := 0; DECLARE v_name VARCHAR(20) := ''; DECLARE v_pay INT := 0; DECLARE v_msg VARCHAR(200) := ''; DELETE FROM message_box; FOR v_id IN 100 .. 103 DO SELECT name, payment INTO v_name, v_pay FROM control_tab WHERE id = :v_id; v_msg := :v_name || ' has ' || TO_CHAR(:v_pay); INSERT INTO message_box VALUES (:v_msg, CURRENT_TIMESTAMP); END FOR; END;

handler Examples

CREATE TABLE TARGET ( A INT); CREATE TABLE MYTAB (I INTEGER PRIMARY KEY); CREATE TRIGGER MYTRIG_SQLEXCEPTION AFTER INSERT ON TARGET BEGIN DECLARE EXIT HANDLER FOR SQLEXCEPTION SELECT ::SQL_ERROR_CODE, ::SQL_ERROR_MESSAGE FROM DUMMY; INSERT INTO MYTAB VALUES (1); INSERT INTO MYTAB VALUES (1); -- expected unique violation error: 301 -- not reached END; CREATE TRIGGER MYTRIG_SQL_ERROR_CODE AFTER UPDATE ON TARGET BEGIN DECLARE EXIT HANDLER FOR SQL_ERROR_CODE 301 SELECT ::SQL_ERROR_CODE, ::SQL_ERROR_MESSAGE FROM DUMMY; INSERT INTO MYTAB VALUES (1); INSERT INTO MYTAB VALUES (1); -- expected unique violation error: 301 -- not reached END; CREATE TRIGGER MYTRIG_CONDITION AFTER DELETE ON TARGET BEGIN DECLARE MYCOND CONDITION FOR SQL_ERROR_CODE 301; DECLARE EXIT HANDLER FOR MYCOND SELECT ::SQL_ERROR_CODE, ::SQL_ERROR_MESSAGE FROM DUMMY; INSERT INTO MYTAB VALUES (1); INSERT INTO MYTAB VALUES (1); -- expected unique violation error: 301 -- not reached END;

SIGNAL/RESIGNAL Examples

CREATE TABLE TARGET ( A INT); CREATE TABLE MYTAB (I INTEGER PRIMARY KEY); CREATE TABLE MYTAB_TRIGGER_ERR (err_code INTEGER, err_msg VARCHAR(30));


100 2012-12-19

CREATE TRIGGER MYTRIG_SIGNAL AFTER INSERT ON TARGET BEGIN DECLARE MYCOND CONDITION FOR SQL_ERROR_CODE 10001; DECLARE EXIT HANDLER FOR MYCOND INSERT INTO MYTAB_TRIGGER_ERR VALUES (::SQL_ERROR_CODE, ::SQL_ERROR_MESSAGE); INSERT INTO MYTAB VALUES (1); SIGNAL MYCOND SET MESSAGE_TEXT = 'my error'; -- not reached END; CREATE TRIGGER MYTRIG_RESIGNAL AFTER UPDATE ON TARGET BEGIN DECLARE MYCOND CONDITION FOR SQL_ERROR_CODE 10001; DECLARE EXIT HANDLER FOR MYCOND RESIGNAL; INSERT INTO MYTAB VALUES (1); SIGNAL MYCOND SET MESSAGE_TEXT = 'my error'; -- not reached END;

Transition variable Example

CREATE TABLE TARGET ( A INT, B VARCHAR(10)); CREATE TABLE SAMPLE_OLD ( A INT, B VARCHAR(10)); CREATE TABLE SAMPLE_NEW ( A INT, B VARCHAR(10)); INSERT INTO TARGET VALUES ( 1, 'oldvalue');

CREATE TRIGGER TEST_TRIGGER_VAR_UPDATE AFTER UPDATE ON TARGET REFERENCING NEW ROW mynewrow, OLD ROW myoldrow FOR EACH ROW BEGIN INSERT INTO SAMPLE_new VALUES(:mynewrow.a, :mynewrow.b); INSERT INTO SAMPLE_old VALUES(:myoldrow.a, :myoldrow.b); END;

UPDATE TARGET SET b = 'newvalue' WHERE A = 1; SELECT * FROM SAMPLE_NEW; 1, 'newvalue' SELECT * FROM SAMPLE_OLD; 1, 'oldvalue'

CREATE VIEW

Syntax

CREATE VIEW <view_name> [(<column_name>, ... )] AS <subquery>

Syntax Elements

<view_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier> <column_name> ::= <identifier>

Description

The CREATE VIEW statement effectively creates virtual table based on the results of an SQL statement. It isnot a table in a real sense as it does not contain data in itself.

When a column name is specified along with the view name, a query result is displayed with that columnname. If a column name is omitted, a query result gives an appropriate name to the column automatically.The number of column names has to be the same as the number of columns returned from <subquery>.

Update operations on views are supported if the following conditions are met:. Each column in the view must map to a column of a single table. If a column in the base table has NOT NULL constraint without default value, the column must be included inview columns to be an insertable view. Update operation on a view is allowed without this condition.. Must not contain an aggregate or analytic function in a SELECT list for example, the followings are notallowed:. TOP, SET, DISTINCT operator in a SELECT list. GROUP BY, ORDER BY clause. Must not contain a subquery in a SELECT list. Must not contain a sequence value(CURRVAL, NEXTVAL)


101 2012-12-19

. Must not contain a column view as the base view

If base views or tables are updatable, a view on the base views or tables can be updatable if the aboveconditions are met.

Example

You create table A.


You create a view v that selects all records from table A.

CREATE VIEW v AS SELECT * FROM A;

DROP AUDIT POLICY

Syntax

DROP AUDIT POLICY <policy_name>

Syntax Elements


Description

The DROP AUDIT POLICY statement drops an audit policy. <policy_name> must specify an existing auditpolicy. Only database users having the system privilege AUDIT ADMIN are allowed to drop an audit policy. Eachdatabase user having this privilege is allowed to drop any audit policy, no matter if he created it or not. Even if an audit policy is dropped, it may happen, that the audit action specified in the dropped audit policywill be audited further. This will happen, if another audit policy is enabled and specifies that audit action aswell. To switch off an audit policy temporarily, it can be disabled and need not be dropped.


AUDIT_POLICY: shows all audit policies and their states M_INIFILE_CONTENTS: shows the configuration parameter concerning auditing

Only database users with system privilege CATALOG READ, DATA ADMIN or INIFILE ADMIN can see any infoout of the view M_INIFILE_CONTENTS. For other database users this view will be empty.

Example

Assuming that the creation of an audit policy happened before using the following statement:


Now this audit policy has to be dropped.

DROP AUDIT POLICY priv_audit;

DROP FULLTEXT INDEX

Syntax

DROP FULLTEXT INDEX <fulltext_index_name>

Syntax Elements

<fulltext_index_name> ::= <identifier>

Specifies the index to be dropped.


102 2012-12-19

Description

The DROP FULLTEXT INDEX statement removes a fulltext index.

Example

You create table A and then add an index i on column b of table A.


You drop index i.

DROP FULLTEXT INDEX i;

DROP INDEX

Syntax

DROP INDEX <index_name>

Syntax Elements

index_name ::= <identifier>

The index name identifier specifies the index to be dropped.

Description

The DROP INDEX statement removes an index.

Example

You create table A and then add an index i on column b of table A.


You drop index i.

DROP INDEX i;

DROP SCHEMA

Syntax

DROP SCHEMA <schema_name> [<drop_option>]

Syntax Elements

<drop_option> ::= CASCADE | RESTRICT

When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedschema, dependent objects of the schema will be invalidated but not dropped.

The invalidated object can be revalidated when an object that has same schema name is created.

CASCADE

Drops the schema and dependent objects.

RESTRICT

Drops the schema only when dependent objects do not exist. If this drop option is used and a dependentobject exists an error will be thrown.

Description


103 2012-12-19

The DROP SCHEMA statement removes a schema.

Example

You create schema my_schema and a table my_schema.t.

CREATE SCHEMA my_schema; CREATE TABLE my_schema.t (a INT);

You drop my_schema with a CASCADE option.

DROP SCHEMA my_schema CASCADE;

DROP SEQUENCE

Syntax

DROP SEQUENCE <sequence_name> [<drop_option>]

Syntax Elements


The name of the sequence to be dropped, with optional schema name.


When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedsequence, dependent objects of the sequence will be invalidated but not dropped.

The invalidated objects can be revalidated when an object that has same schema and object name is created.

CASCADE

Drops the sequence and dependent objects.

RESTRICT

Drops the sequence only when dependent objects do not exist. If this drop option is used and a dependentobject exists an error will be thrown.

Description

The DROP SEQUENCE statement removes a sequence.

Example

You create a sequence seq.

CREATE SEQUENCE seq START WITH 11;

You drop the sequence seq.

DROP SEQUENCE seq;

DROP SYNONYM

Syntax

DROP [PUBLIC] SYNONYM <synonym_name> [<drop_option>]

Syntax Elements

PUBLIC

When specified allows the removal of a public synonym.


104 2012-12-19

<synonym_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The name of the synonym to be dropped, with optional schema name.


When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedsynonym, dependent objects of the synonym will be invalidated but not dropped.

The invalidated object can be revalidated when an object that has same schema and object name is created.

CASCADE

Drops the synonym and dependent objects.

RESTRICT

Drops the synonym only when dependent objects do not exist. If this drop option is used and a dependentobject exists an error will be thrown.

Description

The DROP SYNONYM statement removes a synonym.

Example

You create a table 'a', then a synonym a_synonym and a public synonym pa_synonym for the table.

CREATE TABLE a (c INT); CREATE SYNONYM a_synonym FOR a; CREATE PUBLIC SYNONYM pa_synonym FOR a;

You drop both synonym a_synonym and public synonym pa_synonym using a non-cascaded drop in both cases.

DROP SYNONYM a_synonym; DROP PUBLIC SYNONYM pa_synonym;

DROP TABLE

Syntax

DROP TABLE <table_name> [<drop_option>]

Syntax Elements


The name of the table to be dropped, with optional schema name.


When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedtable, dependent objects of the table will be invalidated but not dropped.


CASCADE

Drops the table and dependent objects.

RESTRICT

Drops the table only when dependent objects do not exist. If this drop option is used and a dependent objectexists an error will be thrown.

Description


105 2012-12-19

The DROP TABLE statement removes a table from the database.

Example

You create table A and then drop it directly afterwards.

CREATE TABLE A(C INT); DROP TABLE A;

DROP TRIGGER

Syntax

DROP TRIGGER <trigger_name>

Syntax Elements

<trigger_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The name of the trigger to be dropped, with optional schema name.


When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedtrigger, dependent objects of the trigger will be invalidated but not dropped.

The invalidated object can be revalidated when an object that has same schema and object name is created.

CASCADE

Drops the trigger and dependent objects.

RESTRICT

Drops the trigger only when dependent objects do not exist. If this drop option is used and a dependentobject exists an error will be thrown.<b

Description

The DROP TRIGGER statement deletes a trigger. Only database users having the TRIGGER privilege for the table on which the trigger was defined are allowedto drop a trigger for that table.

Example

You create two tables, target and sample, and a trigger called test.

CREATE TABLE target ( A INT); CREATE TABLE sample ( A INT); CREATE TRIGGER test AFTER UPDATE ON target BEGIN INSERT INTO sample VALUES(3); END;

You drop the trigger called test.

DROP TRIGGER test;

DROP VIEW

Syntax

DROP VIEW <view_name> [<drop_option>]

Syntax Elements


106 2012-12-19

<view_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The name of the view to be dropped, with optional schema name.


When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedview, dependent objects of the view will be invalidated but not dropped.


CASCADE

Drops the view and dependent objects.

RESTRICT

Drops the view only when dependent objects do not exist. If this drop option is used and a dependent objectexists an error will be thrown.

Description

The DROP VIEW statement removes a view from the database.

Example

You create table t, then a view v that selects all records from table t.

CREATE TABLE t (a INT); CREATE VIEW v AS SELECT * FROM t;

You drop view v from the database.

DROP VIEW v;

RENAME COLUMN

Syntax

RENAME COLUMN <table_name>.<old_column_name> TO <new_column_name>

Syntax Elements

<table_name> ::= <identifier>

The name of the table where the column is to be renamed.

<old_column_name> ::= <identifier> <new_column_name> ::= <identifier>

The old and new column names respectively.

Description

The RENAME COLUMN statement changes the name of a column.

For the information on column name, please refer to Identifiers.

Example

You create table tab with two columns named A and B.

CREATE TABLE tab (A INT PRIMARY KEY, B INT);

You display the column names for table tab stored in the SAP HANA database.

SELECT COLUMN_NAME, POSITION FROM TABLE_COLUMNS WHERE SCHEMA_NAME = CURRENT_SCHEMA


107 2012-12-19

AND TABLE_NAME = 'tab' ORDER BY POSITION;

You rename column A to C.

RENAME COLUMN tab.A TO C;

You display the column names for table tab after the renaming.

SELECT COLUMN_NAME, POSITION FROM TABLE_COLUMNS WHERE SCHEMA_NAME = CURRENT_SCHEMA AND TABLE_NAME = 'tab' ORDER BY POSITION;

RENAME INDEX

Syntax

RENAME INDEX <old_index_name> TO <new_index_name>

Syntax Elements

<old_index_name> ::= <identifier> <new_index_name> ::= <identifier>

Description

The RENAME INDEX statement changes the name of an index.

For the information on index name, please refer to Identifiers.

Example

Table B is created, then index idx is created on column B of table B:

CREATE TABLE B (A INT PRIMARY KEY, B INT); CREATE INDEX idx on B(B);

The list of index names in table B is shown:

SELECT INDEX_NAME FROM INDEXES WHERE SCHEMA_NAME = CURRENT_SCHEMA AND TABLE_NAME = 'B';

Index idx is renamed to new_idx:

RENAME INDEX idx TO new_idx;

The list of index names in table B after renaming is shown:

SELECT INDEX_NAME FROM INDEXES WHERE SCHEMA_NAME = CURRENT_SCHEMA AND TABLE_NAME = 'B';

RENAME TABLE

Syntax

RENAME TABLE <old_table_name> TO <new_table_name>

Syntax Elements

<old_table_name> ::= <identifier> <new_table_name> ::= <identifier>

Description

The RENAME TABLE statement changes the name of a table to new_table_name in the same schema.

For the information on table name, please refer to Identifiers.

Example


108 2012-12-19

Table A is created in current schema:


The list of table names in current schema is shown:

SELECT TABLE_NAME FROM TABLES WHERE SCHEMA_NAME = CURRENT_SCHEMA;

Table A renamed to B:

RENAME TABLE A TO B;

The list of table names in current schema after renaming is shown:

SELECT TABLE_NAME FROM TABLES WHERE SCHEMA_NAME = CURRENT_SCHEMA;

Schema mySchema is created, then table mySchema.A is created:

CREATE SCHEMA mySchema; CREATE TABLE mySchema.A (A INT PRIMARY KEY, B INT);

The list of table names in schema mySchema is shown:

SELECT TABLE_NAME FROM TABLES WHERE SCHEMA_NAME = 'MYSCHEMA';

Table mySchema.A is renamed to B:

RENAME TABLE mySchema.A TO B;

The list of table names in schema mySchema after renaming is shown:

SELECT TABLE_NAME FROM TABLES WHERE SCHEMA_NAME = 'MYSCHEMA';

Data Manipulation Statements

DELETE

Syntax

DELETE [HISTORY] FROM <table_name> [WHERE <condition>]

Syntax Elements

HISTORY

Marks the selected records of the history storage of a history-table for deletion.


For descriptions on identifier, please refer to Identifiers.

<condition> ::= <condition> OR <condition> | <condition> AND <condition> | NOT <condition> | ( <condition> ) | <predicate>

For descriptions on predicates, please refer to Predicates.

Description

The DELETE statement deletes records from a table where the <condition> is met. If the WHERE clause isomitted, then DELETE removes all records from a table.

When using the DELETE HISTORY command time travel queries referencing the deleted rows may still accessthese rows. In order to physically delete these rows the following statements have to be executed:


109 2012-12-19

ALTER SYSTEM RECLAIM VERSION SPACE; MERGE HISTORY DELTA of <table_name>;

In some cases even the execution of the two statements above may not lead to physical deletion.

To check whether the rows are physically deleted, the following statement can be helpful:

SELECT * FROM <table_name> WHERE <condition> WITH PARAMETERS ('REQUEST_FLAGS'=('ALLCOMMITTED','HISTORYONLY'));

Note:The "WITH PARAMETERS ('REQUEST_FLAGS'= ('ALLCOMMITTED','HISTORYONLY'))" clause may onlybe used for the purpose of validating the result of using the DELETE HISTORY statement.

Example

Example 1 - Standard delete operation.

You create a table and insert some data.

CREATE TABLE T (KEY INT PRIMARY KEY, VAL INT); INSERT INTO T VALUES (1, 1); INSERT INTO T VALUES (2, 2); INSERT INTO T VALUES (3, 3);

You delete from table T where the key column is equal to 1;

DELETE FROM T WHERE KEY = 1;

After executing the above query the contents of table T is as follows:

KEY VAL

2 2

3 3

From this result we can see that one row was deleted.

Example 1 - Standard delete operation.

You first create a history table and insert a value.

CREATE HISTORY COLUMN TABLE HIST(KEY INT PRIMARY KEY, VAL INT); INSERT INTO HIST VALUES (1, 1); COMMIT;

You update the row we inserted at the previous step.

UPDATE HIST SET VAL=42 WHERE KEY=1; COMMIT;

You merge in the table delta storage with the table main storage.

MERGE DELTA OF HIST;

Assuming the merge statement has moved the updated record into the history storage of the history table,you can now delete the updated record history.

DELETE HISTORY FROM HIST WHERE KEY=1;

EXPLAIN PLAN

EXPLAIN PLAN

EXPLAIN PLAN [SET STATEMENT_NAME = <statement_name>] FOR \ref sql_subquery

Syntax Elements

<statement_name> ::= string literal used to identify the name of a specific execution plan in the output table for a given SQL statement.


110 2012-12-19

It is set to NULL if the SET STATEMENT_NAME is not specified.

Description

The EXPLAIN PLAN statement is used to evaluate the execution plan that the SAP HANA database follows toexecute an SQL statement. The result of the evaluation is stored into the EXPLAIN_PLAN_TABLE view forlater user examination.

The SQL statement must be data manipulation statement, thus a schema definition language statementcannot be used with the EXPLAIN PLAN command.

You can obtain SQL plan from EXPLAIN_PLAN_TABLE view. The view is shared by all users. Here is an exampleof reading an SQL plan from the view.

SELECT * FROM EXPLAIN_PLAN_TABLE;

Columns in EXPLAIN_PLAN_TABLE view:

Table 1: Column name and description

Column Name Description

STATEMENT_NAMEThe s tring speci fied as STATEMENT_NAME on executing the EXPLAIN PLAN command. Thisi s used to dis tinguish plans from each other when there are multiple plans in theEXPLAIN_PLAN_TABLE view.

OPERATOR_NAME Name of an operator. Deta i l s are described in the fol lowing s ection.

OPERATOR_DETAILSDeta i l s of an operator. Predicates and expres s ions used by the operator are s hownhere.

SCHEMA_NAME Name of the schema of the accessed table.

TABLE_NAME Name of the accessed table.

TABLE_TYPEType of the accessed table. One of the fol lowing options : COLUMN TABLE, ROW TABLE,MONITORING VIEW, JOIN VIEW, OLAP VIEW, CALCULATION VIEW and HIERARCHY VIEW.

TABLE_SIZE Estimated number of rows in the accessed table

OUTPUT_SIZE Estimated number of rows produced by an operator

SUBTREE_COSTEstimated cost of executing the subtree s tarting from an operator. This va lue i s only forrelative comparison.

OPERATOR_ID ID of an operator unique in a plan. IDs are integers s tarting from 1.

PARENT_OPERATOR_IDOPERATOR_ID of the parent of an operator. The shape of an SQL plan i s a tree and thetopology of the tree can be reconstructed us ing OPERATOR_ID andPARENT_OPERATOR_ID. PARENT_OPERATOR_ID of the root operator i s shown as NULL.

LEVELLevel from the root operator. Level of the root operator i s 1, level of a chi ld of the rootoperator i s 2 and so on. This can be uti l i zed for output indentation.

POSITIONPos i tion in the parent operator. Pos i tion of the fi rs t chi ld i s 1, pos i tion of the secondchi ld i s 2 and so on.

HOST The hostname where an operator was executed

PORT The TCP/IP port used to connect to the host

TIMESTAMP Date and time when the EXPLAIN PLAN command was executed.

CONNECTION_ID ID of the connection where the EXPLAIN PLAN command was executed.

EXECUTION_ENGINE Type of the execution engine where an operator i s executed: COLUMN or ROW

OPERATOR_NAME column in EXPLAIN_PLAN_TABLE view: Table 2. List of column engine operators shownin the OPERATOR_NAME column.

Operator Name Des cription

COLUMNSEARCH

Starting pos i tion of column engine operators . OPERATOR_DETAILS l i s ts projected columns .

LIMIT Operator for l imiting the number of output rows

ORDER BY Operator for sorting output rows

HAVING Operator for fi l tering with predicates on top of grouping and aggregation

GROUP BY Operator for grouping and aggregation

DISTINCT Operator for dupl icate el imination

FILTER Operator for fi l tering with predicates

JOIN Operator for joining input relations

COLUMN TABLE Information about accessed column table

MULTIPROVIDEROperator for producing union-a l l of multiple resul ts having the same grouping andaggregation


111 2012-12-19

Table 3. List of row engine operators shown in the OPERATOR_NAME column.

OperatorName

Description

ROW SEARCH Starting pos i tion of row engine operators . OPERATOR_DETAILS l i s ts projected columns .

LIMIT Operator for l imiting number of output rows

ORDER BY Operator for sorting output rows

HAVING Operator for fi l tering with predicates on top of grouping and aggregation

GROUP BY Operator for grouping and aggregation

MERGEAGGREGATION

Operator for merging the resul ts of multiple para l lel grouping and aggregations

DISTINCT Operator for dupl icate el imination

FILTER Operator for fi l tering with predicates

UNION ALL Operator for producing union-a l l of input relations

MATERIALIZEDUNION ALL

Operator for producing union-a l l of input relations with intermediate resul t materia l i zation

BTREE INDEXJOIN

Operator for joining input relations through B-tree index searches . Join type suffix can beadded. For example, B-tree index join for left outer join i s s hown as BTREE INDEX JOIN (LEFTOUTER). Join without join type suffix means inner join.

CPBTREEINDEX JOIN

Operator for joining input relations through CPB-tree index searches . Join type suffix can beadded.

HASH JOINOperator for joining input relations through probing hash table bui l t on the fly. Join type suffixcan be added.

NESTED LOOPJOIN

Operator for joining input relations through nested looping. Join type suffix can be added.

MIXEDINVERTEDINDEX JOIN

Operator for joining an input relation of row s tore format with a column table without formatconvers ion us ing an inverted index of the column table. Join type suffix can be added.

BTREE INDEXSEARCH

Table access through B-tree index search

CPBTREEINDEX SEARCH

Table access through CPB-tree index search

TABLE SCAN Table access through scanning

AGGR TABLE Operator for aggregating base table di rectly

MONITORSEARCH

Monitoring view access through search

MONITORSCAN

Monitoring view access through scanning

COLUMN SEARCH is a mark for the starting position of column engine operators and ROW SEARCH is a markfor the starting position of row engine operators. In the example below, the intermediate result produced bya COLUMN SEARCH (ID 10) is consumed by a ROW SEARCH (ID 7), and the intermediate result produced bythe ROW SEARCH (ID 7) is consumed by another COLUMN SEARCH (ID 1). The operators below the lowestCOLUMN SEARCH (ID 10) explain how the COLUMN SEARCH (ID 10) is executed. The operators between theROW SEARCH (ID 7) and the COLUMN SEARCH (ID 10) explain how the ROW SEARCH (ID 7) processes theintermediate result produced by the COLUMN SEARCH (ID 10). The operators between the top COLUMNSEARCH (ID 1) and the ROW SEARCH (ID 7) explain how the top COLUMN SEARCH (ID 1) processes theintermediate result produced by the ROW SEARCH (ID 7). Table 4. Operators

OPERATOR_NAME OPERATOR_ID PARENT_OPERATOR_ID LEVEL POSITION

COLUMN SEARCH 1 NULL 1 1

LIMIT 2 1 2 1

ORDER BY 3 2 3 1

GROUP BY 4 3 4 1

JOIN 5 4 5 1

COLUMN TABLE 6 5 6 1

<fontcolor="BLUE">ROW SEARCH</font>

<fontcolor="BLUE">7</font>




BTREE INDEX JOIN 8 7 7 1

BTREE INDEX JOIN 9 8 8 1

<fontcolor="RED">COLUMN SEARCH</font>

<fontcolor="RED">10</font>




FILTER 11 10 10 1


112 2012-12-19

COLUMN TABLE 12 11 11 1

Example of SQL plan explanation Here is an example of SQL plan explanation of a query. The query is fromTPC-H Benchmark. In the example, all tables are located on row store.

DELETE FROM expla in_plan_table WHERE s tatement_name = 'TPC-H Q10';

EXPLAIN PLAN SET STATEMENT_NAME = 'TPC-H Q10' FORSELECT TOP 20 c_custkey, c_name, SUM(l_extendedprice * (1 - l_discount)) AS revenue, c_acctba l , n_name, c_address , c_phone, c_commentFROM cus tomer, orders , l inei tem, nationWHERE c_custkey = o_cus tkey AND l_orderkey = o_orderkey AND o_orderdate >= '1993-10-01' AND o_orderdate < ADD_MONTHS('1993-10-01',3) AND l_returnflag = 'R' AND c_nationkey = n_nationkeyGROUP BYc_custkey,c_name,c_acctba l ,c_phone,n_name,c_address ,c_commentORDER BY revenue DESC;

SELECT operator_name, operator_deta i l s , table_nameFROM expla in_plan_tableWHERE s tatement_name = 'TPC-H Q10';

The following is the plan explanation of the above query.

OPERATOR_NAME OPERATOR_DETAILS TABLE_NAME

ROW SEARCHCUSTOMER.C_CUSTKEY, CUSTOMER.C_NAME, SUM(LINEITEM.L_EXTENDEDPRICE * (1 - LINEITEM.L_DISCOUNT)), CUSTOMER.C_ACCTBAL, NATION.N_NAME, CUSTOMER.C_ADDRESS, CUSTOMER.C_PHONE, CUSTOMER.C_COMMENT

None

LIMIT NUM RECORDS: 20

ORDER BY SUM(LINEITEM.L_EXTENDEDPRICE * (1 - LINEITEM.L_DISCOUNT)) DESC None

MERGE AGGREGATION NUM PARTITIONS: 4 None

GROUP BYGROUPING: NATION.N_NAME, R_CUSTOMER.C_CUSTKEY, AGGREGATION: SUM(LINEITEM.L_EXTENDEDPRICE * (1 - LINEITEM.L_DISCOUNT))

None

CPBTREE INDEX JOININDEX NAME: _SYS_TREE_RS_279_#0_#P0,INDEX CONDITION: ORDERS.O_ORDERKEY = LINEITEM.L_ORDERKEY,INDEX FILTER: 'R' = LINEITEM.L_RETURNFLAG

LINEITEM

BTREE INDEX JOININDEX NAME: _SYS_TREE_RS_285_#0_#P0,INDEX CONDITION: CUSTOMER.C_NATIONKEY = NATION.N_NATIONKEY

NATION

BTREE INDEX JOININDEX NAME: _SYS_TREE_RS_283_#0_#P0,INDEX CONDITION: ORDERS.O_CUSTKEY = CUSTOMER.C_CUSTKEY

CUSTOMER

TABLE SCAN FILTER CONDITION: ORDERS.O_ORDERDATE < '1994-01-01' AND ORDERS.O_ORDERDATE >= '1993-10-01' ORDERS

This means that:1. TABLE SCAN will be executed on ORDERS with the FILTER CONDITION.2. BTREE INDEX JOIN will be executed with the B-tree index of CUSTOMER and the result of the below TABLESCAN.3. BTREE INDEX JOIN will be executed with the B-tree index of NATION and the result of the below BTREEINDEX JOIN.4. CPBTREE INDEX JOIN will be executed with the CPB-tree index of LINEITEM and the result of the belowBTREE INDEX JOIN.


113 2012-12-19

5. GROUP BY will be executed with the result of the below CPBTREE INDEX JOIN, with 4 threads.6. MERGE AGGREGATION will be executed with the result of the GROUP BY below.

INSERT

Syntax

INSERT INTO <table_name> [ <column_list_clause> ] { <value_list_clause> | <subquery> }

Syntax Elements

<table_name> ::= [ <schema_name>. ]<identifier> <schema_name> ::= <identifier>

<column_list_clause> ::= ( <column_name>, ... ) <column_name> ::= <identifier>


<value_list_clause> ::= VALUES ( <expression>, ... )

For descriptions on expressions, please refer to Expressions.

Description

The INSERT statement adds records to a table. A subquery that returns records can be used to insert recordsinto the table. If the subquery does not return any records, then the database will not insert any records intothe table. The column list can be specified with the INSERT statement. The column which isn't included in thecolumn list will be shown as a default value. If the column list is omitted, the database inserts all columns inthe table.

Example

You create table T:

CREATE TABLE T (KEY INT PRIMARY KEY, VAL1 INT, VAL2 NVARCHAR(20));

You insert a row into table T.

INSERT INTO T VALUES (1, 1, 'The first');

KEY VAL1 VAL2

1 1 The fi rs t

You insert a new row into table T using column list to specify which columns should receive the input values.

INSERT INTO T (KEY, VAL2) VALUES (2,3);

KEY VAL1 VAL2

1 1 The fi rs t

2 0 NULL

You insert a row into table T using a subquery.

INSERT INTO T SELECT 3, 3, 'The third' FROM DUMMY;

KEY VAL1 VAL2

1 1 The fi rs t

2 0 3

3 3 The thi rd

LOAD

Syntax

LOAD <table_name> [HISTORY] {DELTA | ALL | (<column_name>, ...)}


114 2012-12-19

Syntax Elements

<table_name> ::= [<schema_name>.]<identifier> <schema_name>::= <identifier>

The name of the table to be loaded into memory, with optional schema name.

HISTORY

You can use the HISTORY option when <table_name> is the name of a history table. The default operation ofthe LOAD command loads the current state of the given table into memory. If you explicitly require the tablehistory to be loaded into memory, a separate LOAD statement with the HISTORY option enabled has to beexecuted. So to load a history completely into memory, two LOAD statements have to be executed, with andwithout HISTORY option enabled.

Note:If you do not explicitly load table history into memory the SAP HANA database will automaticallyload this data upon first access.

DELTA

DELTA specifies that the delta part of a column store table is loaded into memory. As the column store is readoptimized and compressed, deltas are used to optimize insert or update operations.

ALL

ALL current data of the column store table, including its delta, is loaded into memory.

Note:This option will not load the table history.


The name of the column to be loaded into memory.

Description

The LOAD statement explicitly loads column store table data into memory instead of upon first access.A LOAD statement without HISTORY option only loads the current table data and the table delta intomemory, but not the history-delta and history-main.


m_cs_tables shows the current table load status.

Examples

You create column table A.

CREATE COLUMN TABLE A (A INT, B INT);

You load table A into memory.

LOAD A all;

You load the columns A and B of table A into memory.

LOAD A (A,B);

You query the load status of table A using the m_cs_tables monitoring view.

select loaded from m_cs_tables where table_name = 'A'

MERGE DELTA

Syntax

MERGE [HISTORY] DELTA OF <table_name> [PART n] [WITH PARAMETERS (<parameter_list>, ...)]


115 2012-12-19

Syntax Elements

HISTORY

Merges the history-delta storage of a table into the history-main storage for column store temporal tables.

PART

Merge the delta of a specific table partition to the tables main storage.


The table name where the delta merge will occur, with optional schema name.

<parameter_list> ::= <parameter> [{, <parameter>}...] <parameter> ::= <parameter_name> = <parameter_setting> <parameter_name> ::= 'SMART_MERGE' | 'MEMORY_MERGE' <parameter_setting> ::= 'ON' | 'OFF'

Column store-specific options are passed using the "WITH PARAMETERS" clause.

SMART_MERGE

When SMART_MERGE is ON, the database does a smart merge based on merge criteria specified inautomerge section of indexserver configuration. When SMART_MERGE is OFF, smart merging is disabled.

MEMORY_MERGE

When MEMORY_MERGE is ON the SAP HANA database merges the table delta storage in memory only, itwill not be persisted. When MEMORY_MERGE is OFF, memory merging is disabled.

Description

The MERGE DELTA statement merges the column store table delta storage to the tables main storage. As acolumn store table is read optimized and compressed deltas are introduced to optimize insert or updateoperations. All insertions are passed to the delta storage. At a certain point in time the delta changes to atable can be merged into the table main storage.

The UPDATE privilege on the column store table is required for performing a delta merge.

Examples

You create history column table A.

CREATE HISTORY COLUMN TABLE A(c NVARCHAR(1000)) PARTITION BY ROUNDROBIN PARTITIONS 2;

You merge the column store table delta storage to the tables main storage.

MERGE DELTA OF A;

You merge the column store table A using a smart merge.

MERGE DELTA OF A WITH PARAMETERS('SMART_MERGE' = 'ON');

You smart merge the column store table A delta storage to main storage in memory only. This merge will notbe persisted to disk.

MERGE DELTA OF A WITH PARAMETERS('SMART_MERGE' = 'ON', 'MEMORY_MERGE' = 'ON');

You merge table A delta storage of partition 1 to the main storage of partion 1.

MERGE DELTA OF A PART 1;

You merge table A history-delta storage into its history-main storage.

MERGE HISTORY DELTA OF A;


116 2012-12-19

You merge table A partition 1 history-delta storage to its history-main storage.

MERGE HISTORY DELTA OF A PART 1;

REPLACE | UPSERT

Syntax

UPSERT <table_name> [ <column_list_clause> ] { <value_list_clause> [ WHERE <condition> | WITH PRIMARY KEY ] | <subquery> } REPLACE <table_name> [ <column_list_clause> ] { <value_list_clause> [ WHERE <condition> | WITH PRIMARY KEY ] | <subquery> }

Syntax Elements

<table_name> ::= [ <schema_name>. ]<identifier> <schema_name> ::= <identifier>

<column_list_clause> ::= ( <column_name>, ... ) <column_name> ::= <identifier>


<value_list_clause> ::= VALUES ( <expression>, ... )




Description

Descriptions to detailed syntax elements of this SQL Command.

The UPSERT or REPLACE statement without a subquery is similar to the UPDATE statement. The onlydifference is when the WHERE clause condition is false, it adds a new record to the table like the INSERTstatement.

In case of a table which has a PRIMARY KEY, the PRIMARY KEY column must be included in the column list.Columns defined with NOT NULL without default specification have to be included in the column list as well.Other columns are filled with default value or NULL if not specified.

The UPSERT or REPLACE statement with a subquery works like the INSERT statement, except that if an oldrow in the table has the same value as a new row for a PRIMARY KEY, then the old row is changed by valuesof the returned record from a subquery. Unless the table has a PRIMARY KEY, it becomes equivalent toINSERT because there is no index to be used to determine whether or not a new row duplicates another.

The UPSERT or REPLACE statement with a 'WITH PRIMARY KEY' is same as one with the subquery. It worksbased on the PRIMARY KEY.

Example

You create table T.

CREATE TABLE T (KEY INT PRIMARY KEY, VAL INT);

You insert a new value.

UPSERT T VALUES (1, 1);

KEY VAL

1 1

You insert a new value if the condition in the WHERE clause is false or update the current row values ifWHERE evaluates to true.


117 2012-12-19

UPSERT T VALUES (2, 2) WHERE KEY = 2;

KEY VAL

1 1

2 2

You update the table row where KEY is equal to 1.

UPSERT T VALUES (1, 9) WHERE KEY = 1;

KEY VAL

1 9

2 2

You use the "WITH PRIMARY KEY" keyword to update the table using the primary key value in the VALUESclause.

UPSERT T VALUES (1, 8) WITH PRIMARY KEY;

KEY VAL

1 8

2 2

You insert values using a subquery.

UPSERT T SELECT KEY + 2, VAL FROM T;

KEY VAL

1 8

2 2

3 8

4 2

SELECT

Syntax

<select_statement> ::= <subquery> [ <for_update> | <time_travel> ] | ( <subquery> ) [ <for_update> | <time_travel> ] <subquery> ::= <select_clause> <from_clause> [<where_clause>] [<group_by_clause>] [<having_clause>] [{<set_operator> <subquery>, ... }] [<order_by_clause>] [<limit>]

Syntax Elements

SELECT clause

The select clause specifies an output to be returned either to the caller or to an outer select clause if oneexists.

<select_clause> ::= SELECT [TOP <unsigned_integer>] [ ALL | DISTINCT ] <select_list>

The select clause.

TOP <unsigned_integer>

Specifies that that the first <unsigned_integer> records from the SQL statement should be returned.

DISTINCT

Specifies that only one copy of each set of duplicate records selected should be returned.


118 2012-12-19

ALL

Specifies that all rows selected, including all copies of duplicates should be returned. This is the defaultoption.

<select_list> ::= <select_item>[, ...]

The list of columns to be retrieved.

<select_item> ::= [<table_name>.] <asterisk> | <expression> [ AS <column_alias> ]

The details of the column to be retrieved.


The table name which will be the source for the selected columns, with optional schema name.

<column_alias> ::= <identifier>

The column alias used to refer to an expression.

<asterisk> ::= *

You use an asterisk (*) to select all columns from all tables or views listed in the FROM clause. If you provide aschema name or a table name with an asterisk(*), it is used to limit the scope of the result set to the specifiedtable.

FROM clause

The from clause specifies inputs such as tables, views, and subqueries to be used in the select statement.

<from_clause> ::= FROM <table>[{, <table>}...]

The list of tables where the SELECT data is sourced from.

<table> ::= <table_name> [ [AS] <table_alias> ] | <subquery> [ [AS] <table_alias> ] | <joined_table>

The table, subquery or join which will be the data source for the query.

<table_alias> ::= <identifier>

The alias identifier for a table, subquery or joined table.

<joined_table> ::= <table> [<join_type>] JOIN <table> ON <predicate> | <table> CROSS JOIN <table> | <joined_table>

Specifies how tables should be joined.

ON <predicate>

Specifies a join predicate.

CROSS JOIN

Specifies that a cross join shoul be performed.

Note:A cross join produces the cross-product of two tables.

<join_type> ::= INNER | { LEFT | RIGHT | FULL } [OUTER]

Specifies the type of join to be performed:

LEFT defines a left outer join.RIGHT defines a right outer join.FULL indicates a full outer join.


119 2012-12-19

OUTER

Specifies that an outer join should be performed.

WHERE clause

<where_clause> ::= WHERE <condition>

The WHERE clause is used to specify predicates on inputs in the FROM clause,


<predicate> ::= <comparison_predicate>| <range_preciate> | <in_predicate> | <exist_predicate> | <like_predicate> | <null_predicate> <comparison_predicate> ::= <expression> { = | != | <> | > | < | >= | <= } [ ANY | SOME | ALL ] ({<expression_list> | <subquery>}) <range_predicate> ::= <expression> [NOT] BETWEEN <expression> AND <expression> <in_predicate> ::= <expression> [NOT] IN ( { <expression_list> | <subquery> } ) <exist_predicate> ::= [NOT] EXISTS ( <subquery> ) <like_predicate> ::= <expression> [NOT] LIKE <expression> [ESCAPE <expression>] <null_predicate> ::= <expression> IS [NOT] NULL

<expression_list> ::= {<expression>, ... }

GROUP BY clause

<group_by_clause> ::= GROUP BY { <group_by_expression_list> | <grouping_set> }

The GROUP BY clause is used to group the selected rows based on the values in the specified columns.

<group_by_expression_list> ::= { <expression>, ... }

<grouping_set> ::= { GROUPING SETS | ROLLUP | CUBE } [BEST <signed_integer>] [LIMIT <unsigned_integer>[OFFSET <unsigned_integer>] ] [WITH SUBTOTAL] [WITH BALANCE] [WITH TOTAL] [TEXT_FILTER <filterspec> [FILL UP [SORT MATCHES TO TOP]]] [STRUCTURED RESULT [WITH OVERVIEW] [PREFIX <prefix_table_name>] | MULTIPLE RESULTSETS] ( <grouping_expression_list> )

<grouping_expression_list> ::= { <grouping_expression>, ... }

<grouping_expression> ::= <expression> | ( <expression>, ... ) | ( ( <expression>, ... ) <order_by_clause> )

GROUPING SETS

Generates results with specified multiple groupings of data in a single statement. If no additional options,such as BEST and LIMIT, are set the result produced will be the same as a UNION ALL of the aggregation ofeach specified group. For example

select col1, col2, col3, * count(*) from t group by grouping sets ( (col1, col2), (col1, col3) ) \encode is equivalent to \code select * col1, col2, NULL, count(*) from t group by col1, col2 union all select col1, NULL, * col3, count(*) from t group by col1, col3

In the grouping-sets query each of (col1, col2) and (col1, col3) specifies the grouping.

ROLLUP


120 2012-12-19

Generates results with multiple levels of aggregation in a single statement.For example

rollup (col1, col2, col3)

is equivalent to

grouping sets ( (col1, col2, col3), (col1, col2), (col1) )

with an additional aggregation without grouping. Thus, the number of grouping sets that result set containsis the number of columns in ROLLUP list plus one for a last aggregation if there is no additional option.

CUBE

Generates results with multiple levels of aggregations in a single statement.For example

cube (col1, col2, col3)

is equivalent to

grouping sets ( (col1, col2, col3), (col1, col2), (col1, col3), (col2, col3), (col1), (col2), (col3) )

with an additional aggregation without grouping. Thus, the number of grouping sets that result set containsis the same as all possible permutations of columns in the CUBE list plus one for the last aggregation if there isno additional option.

BEST <signed_integer>

Returns only the top-n grouping sets sorted in descending order of the number of rows aggregated in eachgrouping set. <signed_integer> can be zero, positive, and negative. When n is zero, it is the same as the BESToption not being set. When n is negative, it means sorting in ascending order.

LIMIT <unsigned_integer> [OFFSET <unsigned_integer>]

Returns the first <unsigned_integer> grouped records after skipping OFFSET<unsigned_integer> for eachgrouping set.

WITH SUBTOTAL

Returns an additional subtotal for each grouping set in the returned results controlled by OFFSET or LIMIT.Unless OFFSET and LIMIT are specified, the value is the same as WITH TOTAL.

WITH BALANCE

Returns for each grouping set an additional aggregated value of the remaining values not returned byOFFSET or LIMIT.

WITH TOTAL

Returns for each grouping set an additional row that is the aggregated total value. OFFSET and LIMIT optionsdo not change this value.

TEXT_FILTER <filterspec>

Performs text filtering or highlighting on the grouping columns using a <filterspec>, which is a single-quotedstring that follows the below syntax.

<filterspec> ::= '[<prefix>]<element>{<subsequent>, ...}' <prefix> ::= + | - | NOT <element> ::= <token> | <phrase> <token> ::= !! Unicode letters or digits <phrase> ::= !! double-quoted string that does not contain double quotations inside <subsequent> ::= [<prefix_subsequent>]<element> <prefix_subsequent> ::= + | - | NOT | AND | AND NOT | OR

A filter defined by <filterspec> is a token/phrase or tokens/phrases connected with logical operators such asAND, OR and NOT. A token matches a string that contains its corresponding word in a case-insensitive


121 2012-12-19

manner. For example, 'ab' matches 'ab cd' and 'cd Ab' but does not match 'abcd'. A token can contain thewildcard character ‘*’ that matches any string and ‘?’ that matches any character. Inside a phrase ‘*’ and ‘?’do not function as wildcard characters. With tokens and phrases logical operators AND, OR and NOT can beused together. As OR is the default operator, 'ab cd' is the same as 'ab OR cd'

Note:Logical operators should be written in uppercase characters.

Prefixes '+' and '-' mean inclusion (AND) and exclusion (AND NOT), respectively. For example, 'ab -cd' is thesame as 'ab AND NOT cd'. If FILL UP is not specified, only grouped records that have matching values arereturned.

Note:A filter is applied only to the first grouping column in each grouping set.

FILL UP

Returns not only matched grouped records, but also non-matched records. TEXT_FILTER is useful to identifywhich records have been matched. See 'Related Functions' below.

SORT MATCHES TO TOP

Returns matching values before non-matching values for each grouping set. This option cannot be used withSUBTOTAL, BALANCE and TOTAL.

STRUCTURED RESULT

Returns results as temporary tables. For each grouping set a single temporary table is created. If the WITHOVERVIEW option is set, an additional temporary table is created for the overview of the grouping sets. Thenames of the temporary tables are specified by the PREFIX option.

WITH OVERVIEW

Returns an overview in a separate additional table.

PREFIX <prefix_table_name> <prefix_table_name> ::= #<identifier>

Specifies a prefix for naming the temporary tables. <prefix_table_name> must start with "#", which means atemporary table. If PREFIX is not specified, the default prefix is "#GN" followed by a nonnegative integernumber. See 'Return Format' below.

MULTIPLE RESULTSETS

Specifies that results should be returned in multiple result sets.

Related Functions

grouping_id ( <grouping_column1, ..., grouping_columnn> ) function returns an integer number toidentify which grouping set each grouped record belongs to.text_filter ( <grouping_column> ) function, which is used with TEXT_FILTER, FILL UP, and SORTMATCHES TO TOP, displays matching values or NULL. NULL is displayed for non-matching values whenFILL UP option is specified.

Return Format

If neither STRUCTURED RESULT nor MULTIPLE RESULTSETS is set, the unioned result ofall groupingsets is returned, with NULL values filling up attributes that are not included in a specific grouping set.With STRUCTURED RESULT, temporary tables are created additionally which can be queried using"SELECT * FROM <table name>" in the same session. The name of the tables follows the form:

<PREFIX>0: this table will contain the overview if WITH OVERVIEW is specified <PREFIX>n: n-th grouping set subject to re-ordering by the BEST parameter

With MULTIPLE RESULTSETS, multiple result sets are returned. Grouped records for each grouping setare in a single result set.

HAVING clause

<having_clause> ::= HAVING <condition>

Selects the specified groups that satisfy the predicates. If this clause is omitted, all groups are selected.


122 2012-12-19

SET OPERATORS

The set operators enable multiple select statements to be combined but return only one result set.

<set_operator> ::= UNION [ ALL | DISTINCT ] | INTERSECT [DISTINCT] | EXCEPT [DISTINCT]

UNION ALL

Selects all records from all select statements. Duplicates are not removed.

UNION [DISTINCT]

Selects all unique records from all select statements by removing duplicates found from different selectstatements. UNION has the same function as UNION DISTINCT.

INTERSECT [DISTINCT]

Returns records that exist in all select statement results.

EXCEPT [DISTINCT]

Returns all unique records from the first select statements after removing the duplicates in the followingselect statements.

ORDER BY clause

<order_by_clause> ::= ORDER BY { <order_by_expression>, ... }

The ORDER BY clause is used to sort records by expressions or positions.

<order_by_expression> ::= <expression> [ ASC | DESC ] [ NULLS FIRST | NULLS LAST ] | <position> [ ASC | DESC ] [ NULLS FIRST | NULLS LAST ] <position> ::= <unsigned_integer>

<position> uses the entries in the select list to define the ordering required. For example:

SELECT col1, col2 FROM t ORDER BY 2

"ORDER BY 2" indicates that ordering should be undertaken using the second expression in the select list,which in this case is "col2".

[ ASC | DESC ] \encode Default: ASC ASC sorts records in ascending order. DESC sorts records in descending order.<br>

\code [ NULLS FIRST | NULLS LAST ]

Specifies where in the results set NULL values should appear. By default for ascending ordering NULL valuesare returned first, and for descending they are returned last. You can override this behavior using NULLSFIRST or NULLS LAST to explicitly specify NULL value ordering.

LIMIT

<limit> ::= LIMIT <unsigned_integer> [ OFFSET <unsigned_integer> ]

Limits the number of output records. For example:

LIMIT 3 [OFFSET 5]

Returns the first 3 records after skipping 5 records.

FOR UPDATE

<for_update> ::= FOR UPDATE [OF <update_column_name_list>]

Locks the selected records so that other users cannot lock or change the records until end of this transaction.


123 2012-12-19

<update_column_name_list>> ::= ( <column_name>[{, <column_name>}...] )

A list of columns is used to specify which table or view in the FROM clause should be locked. If there is only asingle table or view, FOR UPDATE will suffice. However, if there are two objects (e.g., tables, views, and,subqueries), OF should be used.

Note:A table used in a subquery cannot be locked. Also that only one table/view can be currently locked.

The lock is released when the corresponding transaction is finished by commit or rollback.

create column table x ( a int, b int ); insert into x values (1,1); insert into x values (2,2); create column table y ( a int, b int ); insert into y values (1,1); insert into y values (2,2); select * from x where a=1 for update; --> OK commit; select * from x where a=1 for update of x.c; --> error because c column does not exist in table x. select * from x where a=1 for update of x.a; --> OK commit; select * from x, y where a=1 for update; --> error because "OF" is not used to specify which one of x and y is locked. select * from x, y where a=1 for update of y.b; --> OK commit; select * from x, (select * from y) where a=1 for update of y.b; --> error because y is inside subquery commit;

TIME TRAVEL

Keywords related with time travel can be used for statement-level time travel to go back to the snapshotspecified by commit_id or utctimestamp.

<time_travel> ::= AS OF { { COMMIT ID <commit_id> } | { UTCTIMESTAMP <timestamp> } }

Time travel is possible only with a history column table. <commit_id> can be obtained fromm_history_index_last_commit_id after each commit and its related <timestamp> can be obtained fromsys.m_transaction_history.

For example:

create history column table x ( a int, b int ); // after turnning off auto commit insert into x values (1,1); commit; select last_commit_id from m_history_index_last_commit_id where session_id = current_connection; // e.g., 10 insert into x values (2,2); commit; select last_commit_id from m_history_index_last_commit_id where session_id = current_connection; // e.g., 20 delete from x; commit; select last_commit_id from m_history_index_last_commit_id where session_id = current_connection; // e.g., 30 select * from x as of commit id 30; // return nothing select * from x as of commit id 20; // return two records (1,1) and (2,2) select * from x as of commit id 10; // return one record (1,1)

select commit_time from sys.transaction_history where commit_id = 10; // e.g., '2012-01-01 01:11:11' select commit_time from sys.transaction_history where commit_id = 20; // e.g., '2012-01-01 02:22:22' select commit_time from sys.transaction_history where commit_id = 30; // e.g., '2012-01-01 03:33:33'

select * from x as of utctimestamp '2012-01-02 02:00:00'; // return one record (1,1) select * from x as of utctimestamp '2012-01-03 03:00:00'; // return two records (1,1) and (2,2) select * from x as of utctimestamp '2012-01-04 04:00:00'; // return nothing


124 2012-12-19

Examples

You create column table t1 and populate it with some example data.

create column table t1 ( id int primary key, customer varchar(5), year int, product varchar(5), sales int ); insert into t1 values(1, 'C1', 2009, 'P1', 100); insert into t1 values(2, 'C1', 2009, 'P2', 200); insert into t1 values(3, 'C1', 2010, 'P1', 50); insert into t1 values(4, 'C1', 2010, 'P2', 150); insert into t1 values(5, 'C2', 2009, 'P1', 200); insert into t1 values(6, 'C2', 2009, 'P2', 300); insert into t1 values(7, 'C2', 2010, 'P1', 100); insert into t1 values(8, 'C2', 2010, 'P2', 150);

Example 1 - GROUPING SETS

You use grouping sets to analyze the customer data.

Note:The this query is equivalent to example 3 below.

select customer, year, product, sum(sales) from t1 group by GROUPING SETS ( (customer, year), (customer, product) );

select customer, year, NULL, sum(sales) from t1 group by customer, year union all select customer, NULL, product, sum(sales) from t1 group by customer, product;

Observe that the two groups inside grouping sets in the first query are specified in each GROUP BY clause inthe second query.

Example 2 - ROLLUP

You use ROLLUP to generates results with multiple levels of aggregation.

Note:The following ROLLUP query is equivalent to Example 4 below.

select customer, year, sum(sales) from t1 group by ROLLUP(customer, year);

select customer, year, sum(sales) from t1 group by grouping sets ( (customer, year), (customer) ) union all select NULL, NULL, sum(sales) from t1;

Example 3 - CUBE

You use CUBE to generates results with multiple levels of aggregation.

Note:The following CUBE query is equivalent to the query in Example 5.

select customer, year, sum(sales) from t1 group by CUBE(customer, year);


125 2012-12-19

select customer, year, sum(sales) from t1 group by grouping sets ( (customer, year), (customer), (year) ) union all select NULL, NULL, sum(sales) from t1;

Example 4 - BEST 1

You use BEST 1 to return only the best group of results.

select customer, year, product, sum(sales) from t1 group by grouping sets BEST 1 ( (customer, year), (product) );

In this example, 4 records exist for (customer, year) group and 2 records exist for (product) group, so theformer 4 records are returned. If you used 'BEST -1' instead of 'BEST 1', the latter 2 records would bereturned.

Example 5 - LIMIT

You use LIMIT 2 to limit the number of records to a maximum 2 for each group.

select customer, year, product, sum(sales) from t1 group by grouping sets LIMIT 2 ( (customer, year), (product) );

For the (customer, year) group, the number of records are 4, so only first 2 records will be returned. For the(product) group, the number of records are 2, in this case all the records will be returned.

Example 6 - WITH SUBTOTAL

you use WITH SUBTOTAL to produce an additional record for each group that displays a subtotal of returnedrecords. These subtotal records are NULL for the customer, year, product columns and the sum of sum(sales)values in the select list.

select customer, year, product, sum(sales) from t1 group by grouping sets LIMIT 2 WITH SUBTOTAL ( (customer, year), (product) );

Example 7 - WITH BALANCE

You use WITH BALANCE to produce an additional record for each group that displays a subtotal of unreturnedrecords.

select customer, year, product, sum(sales) from t1 group by grouping sets LIMIT 2 WITH BALANCE ( (customer, year), (product) );

Example 8 - WITH TOTAL

You use WITH TOTAL to produce an additional record for each group that displays a total of all grouped


126 2012-12-19

records.

select customer, year, product, sum(sales) from t1 group by grouping sets LIMIT 2 WITH TOTAL ( (customer, year), (product) );

Example 9 - TEXT_FILTER

You use TEXT_FILTER to retrieve the first column of each group with a given <filterspec>. The following querywill search for columns ending with '2'. This will be applied to customers for in first grouping set and productsin the second. Only three matched records will be returned.

select customer, year, product, sum(sales), text_filter(customer), text_filter(product) from t1 group by grouping sets TEXT_FILTER '*2' ( (customer, year), (product) );

Example 10 - FILL UP

You use FILL UP to return both matched and non-matched records with <filterspec>. The following queryreturns six records whereas the previous example only returned three.

select customer, year, product, sum(sales), text_filter(customer), text_filter(product) from t1 group by grouping sets TEXT_FILTER '*2' FILL UP ( (customer, year), (product) );

Example 11 - SORT MATCHES TO TOP

SORT MATCHES TO TOP is used to raise matched records up. For each grouping set, its grouped records willbe sorted.

select customer, year, product, sum(sales), text_filter(customer), text_filter(product) from t1 group by grouping sets TEXT_FILTER '*2' FILL UP SORT MATCHES TO TOP ( (customer, year), (product) );

Example 12 - STRUCTURED RESULT

You use STRUCTURED RESULT to create temporary tables, one for each grouping set and an additional tablefor the overview table.

select customer, year, product, sum(sales) from t1 group by grouping sets STRUCTURED RESULT ( (customer, year), (product) );

select * from "#GN1"; select * from "#GN2";

"#GN1" table is for (customer, year) grouping set and "#GN2" table is for (product) one.

Note:Each table contains only related columns. That is to say, "#GN1" table does not have "product"


127 2012-12-19

column and "#GN2" table does not have "customer" and "year" columns.

You use WITH OVERVIEW to create a temporary table "#GN0" for the overview table.

drop table "#G1"; drop table "#G2";

select customer, year, product, sum(sales) from t1 group by grouping sets structured result WITH OVERVIEW ( (customer, year), (product) );

select * from "#GN0"; select * from "#GN1"; select * from "#GN2";

You change the names of temporary tables by using the PREFIX keyword.

Note:the names of tables must still must start with '#' which is the prefix used for temporary tables.

select customer, year, product, sum(sales) from t1 group by grouping sets STRUCTURED RESULT WITH OVERVIEW PREFIX '#MYTAB' ( (customer, year), (product) );

select * from "#MYTAB0"; select * from "#MYTAB1"; select * from "#MYTAB2";

Temporary tables are dropped when the corresponding session is closed or when a user executes a dropcommand. A list of temporary tables are seen in m_temporary_tables.

select * from m_temporary_tables;

Example 13 - MULTIPLE RESULTSETS

You use MULTIPLE RESULTSETS to return multiple result sets. The following query will return three resultsets, one for the overview table and two for the grouping sets.

select customer, year, product, sum(sales) from t1 group by grouping sets MULTIPLE RESULTSETS ( (customer, year), (product) );

TRUNCATE TABLE

Syntax

TRUNCATE TABLE <table_name>

Description

Deletes all rows from a table. TRUNCATE is faster than DELETE FROM when deleting all records from a table,but TRUNCATE cannot be rolled back. To rollback from deleting records, "DELETE FROM <table_name>"should be used.

HISTORY tables can also be truncated in the same way as normal tables by executing this command. All partsof the history table (main, delta, history main and history delta) will be deleted and the content will be lost.

Example

You create table A.


128 2012-12-19


You truncate the contents of table A.

TRUNCATE TABLE A

UNLOAD

Syntax

UNLOAD <table_name>

Syntax Elements


The name of the table to be unloaded from memory

Description

The UNLOAD statement unloads the column store table from memory. This can be done to free up memory.The table will be loaded again on next access.

Example

You create column table A.

CREATE COLUMN TABLE A (A INT, B INT);

You load table A into memory.

LOAD A all;

You unload table A from memory.

UNLOAD A;

You check the load status of table A.

select loaded from m_cs_tables where table_name = 'A'

UPDATE

Syntax

UPDATE <target_table> <set_clause> [ <from_clause> ] [ WHERE <condition> ]

Syntax Elements

<target_table> ::= <table_name> | <alias_name> <table_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier> <alias_name> ::= [AS] <identifier>


<set_clause> ::= SET {<column_name> = <expression>},...


<from_clause> ::= FROM {<table>, ...}

For detailed descriptions on from_clause, please refer to FROM clause.

<condition> ::= <condition> OR <condition> | <condition> AND <condition>


129 2012-12-19

| NOT <condition> | ( <condition> ) | <predicate>


Description

The UPDATE statement changes the values of the records of a table where the condition is met. The FROMclause can specify that a table, view, or derived object that is used to provide the criteria for the updateoperation. If the FROM clause exists and the object being updated is included in the FROM clause, there isonly one reference to the object being updated. If the WHERE clause condition is true, the result ofexpression is assigned to that column. If the WHERE clause is omitted, then it updates all records of a table.

Examples

You create table T, and insert two rows into it.

CREATE TABLE T (KEY INT PRIMARY KEY, VAL INT); INSERT INTO T VALUES (1, 1); INSERT INTO T VALUES (2, 2);

You update the rows of table T if the condition in the WHERE clause is true.

UPDATE T SET VAL = VAL + 1 WHERE KEY = 1;

KEY VAL

1 2

2 2

You update all rows of table T because a where clause is not specified as part of the update statement.

UPDATE T SET VAL = KEY + 10;

KEY VAL

1 11

2 12

You create table T2, and insert two rows into it.

CREATE TABLE T2 (KEY INT PRIMARY KEY, VAR INT); INSERT INTO T2 VALUES (1, 2); INSERT INTO T2 VALUES (3, 6);

You update the values of table T by joining the target table T with table T2.

UPDATE T SET VAL = T2.VAR FROM T, T2 WHERE T.KEY = T2.KEY;

KEY VAL

1 2

2 2

System Management Statements

ALTER SYSTEM ALTER CONFIGURATION

Syntax

ALTER CONFIGURATION (<filename>, <layer>[, <layer_name>]) SET | UNSET <parameter_key_value_list> [ WITH RECONFIGURE]

Syntax Elements

<filename> ::= <string_literal>

The filename is 'indexserver.ini' in the case of row-store engine configuration. The filename used must be oneof the ini files located on the 'DEFAULT' layer. If the file selected by filename does not exist on the requiredlayer, the file will be created in the case of a SET command.


130 2012-12-19

<layer> ::= <string_literal>

Sets the target layer for the configuration change. This parameter can be either 'SYSTEM' or 'HOST'. TheSYSTEM layer is the recommended layer for customer settings. The HOST layer should generally only be usedfor minor configuration, for example parameters contained in daemon.ini.

<layer_name> ::= <string_literal>

If the layer parameter above is set to 'HOST', layer_name is used to target either a tenant name or a targethost name. For example, 'selxeon12' would target the 'selxeon12' host.

SET The SET command updates the value of a key if the key already exists, or inserts it if required.

UNSET The UNSET command removes a key and its associated value.

<parameter_key_value_list> ::= {(<section_name>,<parameter_name>) = <parameter_value>},...

Specifies the section, key and value of the ini file parameter to be altered as follows:

<section_name> ::= <string_literal>

The section name of the parameter to be altered.

<parameter_name> ::= <string_literal>

The name of the parameter to be altered.

<parameter_value> ::= <string_literal>

The value of the parameter to be altered.

WITH RECONFIGURE When WITH RECONFIGURE is specified the configuration changes will be directly applied to the running SAPHANA database instance.When WITH RECONFIGURE is not specified the new configuration will be written to the ini file, however thenew value is not applied to the current running system and will only be applied at the next startup of thedatabase. This means that there can be inconsistencies between the ini file contents and the actualconfiguration value that the SAP HANA database is currently using.

Description

Sets or removes configuration parameters in an ini file. Ini file configuration is used for the layeredconfiguration for DEFAULT, SYSTEM, HOST layers.

Note:The DEFAULT layer configuration cannot be changed or removed using this command.

The following is an example of ini file locations:

DEFAULT: /usr/sap/<SYSTEMNAME>/HDB<INSTANCENUMBER>/exe/config/indexserver.iniSYSTEM: /usr/sap/<SYSTEMNAME>/SYS/global/hdb/custom/config/indexserver.iniHOST: /usr/sap/<SYSTEMNAME>/HDB<INSTANCENUMBER>/<HOSTNAME>/indexserver.ini

The priority of the configuration layers is as follows: DEFAULT < SYSTEM < HOST.

This means that the layer that has the highest priority is the HOST layer, followed by the SYSTEM layer andfinally the DEFAULT layer. The configuration with the highest priority will be applied to the runningenvironment. If the highest priority level configuration is removed, then the configuration with the nexthighest priority will be applied.


Currently available ini files are listed in the M_INIFILES system table and the current configuration is availablein M_INIFILE_CONTENTS system table.

Example

You set a parameter new_test_value in the alt_sys_test section of the golbal.ini file.


131 2012-12-19

ALTER SYSTEM ALTER CONFIGURATION ('global.ini', 'SYSTEM') SET ('alt_sys_test', 'new_test_value') = 'test';

You check the setting of the parameter in the SAP HANA studio. In the Navigator you right click on thesystem where you made the change and select Administration then select the Configuration tab. Expandthe global.ini configuration file and then the alt_sys_test section.

You unset the new_test_value parameter set in the previous step.

ALTER SYSTEM ALTER CONFIGURATION ('global.ini', 'SYSTEM') UNSET ('alt_sys_test', 'new_test_value');

You check the removal of the parameter by clicking refresh on the Configuration page in the SAP HANAstudio.

ALTER SYSTEM ALTER SESSION SET

Syntax

ALTER SYSTEM ALTER SESSION <session_id> SET <key> = <value>

Syntax Elements

<session_id> ::= <unsigned_integer>

The session ID of the session where the variable should be set.

<key> ::= <string_literal>

The key of a session variable. The maximum length of key is 32 characters.

<value> ::= <string_literal>

The desired value of a session variable. The maximum length of value is 512 characters.

Description

With this command you can set session variables of database sessions.

Note:There are several read-only session variables that you cannot change with this command:APPLICATION, APPLICATIONUSER, TRACEPROFILE.

Session variables can be retrieved using SESSION_CONTEXT function and unset using the ALTER SYSTEMALTER SESSION UNSET command.

Example

Note:Please run each command in this example in the same SQL Editor in the SAP HANA studio. This isbecause you need to use the same database session to make the example work properly.

You obtain your current session ID.

SELECT connection_id FROM m_connections WHERE OWN = 'TRUE';

You set the variable MY_VAR to 'some_value' in your session. In the command below replace<your_session_id> with the session ID that you discovered from the command above.

ALTER SYSTEM ALTER SESSION <your_session_id> SET 'MY_VAR'= 'some_value';

You check the current value of MY_VAR in your session.

SELECT SESSION_CONTEXT('MY_VAR') FROM dummy;

ALTER SYSTEM ALTER SESSION UNSET

Syntax


132 2012-12-19

ALTER SYSTEM ALTER SESSION <session_id> UNSET <key>

Syntax Elements

<session_id> ::= <unsigned_integer>

The session ID of the session where the variable should be unset.



Description

With this command you can unset session variables of database sessions.

Note:There are several read-only session variables that you cannot change with this command:APPLICATION, APPLICATIONUSER, TRACEPROFILE.

Session variables can be retrieved using SESSION_CONTEXT function.

Example

You set the session variable MY_VAR to 'abc' in your database session.

SET 'MY_VAR' = 'abc';

You get a list of all the session variables of the current session.

SELECT * FROM M_SESSION_CONTEXT WHERE CONNECTION_ID = CURRENT_CONNECTION

You remove the session variable from the specified session. In the command below replace <your_session_id>with the session ID that you discovered from the command above.

ALTER SYSTEM ALTER SESSION <your_session_id> UNSET 'MY_VAR';

You get a list of all the session variables of the current session.

SELECT * FROM M_SESSION_CONTEXT WHERE CONNECTION_ID = CURRENT_CONNECTION

From the results of this command you can see that the MY_VAR variable has been unset.

ALTER SYSTEM CANCEL [WORK IN] SESSION

Syntax

ALTER SYSTEM CANCEL [WORK IN] SESSION <session_id>

Syntax Elements

<session_id> ::= <string_literal>

The session ID of the required session.

Description

Cancels the currently executing statement of a session. The transaction of the cancelled session will be rolledback. The statement that was executing will return error code 139 (current operation cancelled by requestand transaction rolled back).

Example

You use the following query to obtain current database connection IDs and to view the statements that thesessions are executing.

SELECT C.CONNECTION_ID, PS.STATEMENT_STRING


133 2012-12-19

FROM M_CONNECTIONS C JOIN M_PREPARED_STATEMENTS PS ON C.CONNECTION_ID = PS.CONNECTION_ID AND C.CURRENT_STATEMENT_ID = PS.STATEMENT_ID WHERE C.CONNECTION_STATUS = 'RUNNING' AND C.CONNECTION_TYPE = 'Remote'

Using the connection ID information you obtained using the query above, you can now cancel a runningquery. In the statement below you replace <connection_id> with a connection ID from the query above.

ALTER SYSTEM CANCEL SESSION '<connection_id>'

ALTER SYSTEM CLEAR SQL PLAN CACHE

Syntax


Description

The SQL PLAN CACHE stores plans generated by previous SQL statement executions. The plan cache is used bythe SAP HANA database to speed up query exectution if the same SQL statement is executed again.The plancache also collects some statistics regarding plan preparation and execution.

You can find details about the contents of the SQL plan cache from the following monitoring views:M_SQL_PLAN_CACHE, M_SQL_PLAN_CACHE_OVERVIEW

The ALTER SYSTEM CLEAR SQL PLAN CACHE statement removes all the SQL plans that are not currentlyexecuting from the plan cache. The command also removes all plans having reference count of '0' from theplan cache and resets all the statistics of the remaining plans. Lastly the command also reset the contents ofM_SQL_PLAN_CACHE_OVERVIEW monitoring view.

Example

You clear the SQL plan cache.


ALTER SYSTEM CLEAR TRACES

Syntax

ALTER SYSTEM CLEAR TRACES (<trace_type_list>)

Syntax Elements

<trace_type_list> ::= <trace_type> [,...]

You can clear multiple traces simultaneously by adding multiple trace_types in a comma delimited list.

<trace_type> ::= <string_literal>

You can selectively clear specific trace files by setting the trace_type to one of the following types:

<trace_type> Trace Files

ALERT *a lert_*.trc

CLIENT loca lcl ient_*.trc

CRASHDUMP *.crashdump.*

EMERGENCYDUMP *.emergencydump.*

* a l l *.trc fi les of services l i s ted below

INDEXSERVER,NAMESERVER,...,DAEMON open *.trc fi les of a s ingle service type

Description

You use the ALTER SYSTEM CLEAR TRACES to clear trace contents from trace files. When you use thiscommand all trace files that were opened by the SAP HANA database will be removed or cleared. Ondistributed systems, this command will clear all trace files on all hosts.


134 2012-12-19

distributed systems, this command will clear all trace files on all hosts.

Use this command to reduce disk space used by large tracefiles, e.g. when trace components are set to INFOor DEBUG.

You can monitor trace files and trace file contents using the M_TRACEFILES, M_TRACEFILE_CONTENTSsystem tables respectively.

Examples

You clear the alert trace file.

ALTER SYSTEM CLEAR TRACES('ALERT');

You clear the alert and client trace files.

ALTER SYSTEM CLEAR TRACES('ALERT', 'CLIENT');

ALTER SYSTEM DISCONNECT SESSION

Syntax

ALTER SYSTEM DISCONNECT SESSION <session_id>

Syntax Elements

<session_id> ::= <string_literal>

The session ID of the session to be disconnected.

Description

You use ALTER SYSTEM DISCONNECT SESSION to disconnect a specified session from the database. Beforedisconnection any currently running operations associated with the session will be terminated.

Example

You use the command below to obtain the session IDs of idle sessions.

SELECT CONNECTION_ID, IDLE_TIME FROM M_CONNECTIONS WHERE CONNECTION_STATUS = 'IDLE' AND CONNECTION_TYPE = 'Remote' ORDER BY IDLE_TIME DESC

You disconnect a session. In the statement below you replace <connection_id> with a connection ID from thequery above.

ALTER SYSTEM DISCONNECT SESSION '<connection_id>'

ALTER SYSTEM LOGGING

Syntax

ALTER SYSTEM LOGGING <on_off>

Syntax Elements

<on_off> ::= ON | OFF

ONEnable logging.

OFFDisable logging.

Description

You used this command to Enable or disable logging.

While logging is disabled no log entries will be persisted, only the data area will be written when a savepoint


135 2012-12-19

is reached. This can cause loss of committed transactions, when the indexserver terminated in the middle ofa LOAD operation. In case of a termination, you will have to truncate and insert all data again.

After enabling logging you have to perform a savepoint to be sure that all data will be persisted. You will alsohave to perform a data backup, otherwise you will be unable to recover this data.

ALTER SYSTEM LOGGING waits for the end (commit/rollback) of existing active write transactions. During theexecution this command write transactions are blocked. In the event that there is a long-running writetransaction, this command can be failed with a lock wait timeout error. Starting a new write transaction alsocan be blocked and failed with the same error.

Due to the potential generation of the errors mentioned above you should only use this command for initialload. If used at other times data loss can occur if the indexserver is terminated whilst logging is disabled.Consequently, in a running system, the logging mode should be always ON and should not be altered.

You can also adjust the logging settings for single column tables with ALTER TABLE <table_name> [ENABLE |DISABLE] DELTA LOG

Example

You disable system logging.

ALTER SYSTEM LOGGING OFF;

You enable system logging.

ALTER SYSTEM LOGGING ON;

See Also

ALTER TABLE

ALTER SYSTEM PERSISTENCE ENCRYPTION

Syntax

ALTER SYSTEM PERSISTENCE ENCRYPTION <encrypt_option>

Syntax Elements

<encrypt_option> ::= ON | OFF | CREATE NEW KEY | APPLY CURRENT KEY

ONSwitch on encryption.

OFFSwitch off encryption.

CREATE NEW KEYForce the creation of a new random encryption key.

APPLY CURRENT KEYForce all existing old data to be encrypted using the current encryption key.

Description

You use ALTER SYSTEM PERSISTENCE ENCRYPTION to define if persistent data should be stored on disk inencrypted or non-encrypted format. You can also use this command to recreate the random encryption key.

When you switch on encryption, a random encryption key is prepared and an asynchronous background taskis started which encrypts all disk data with this key.

When you switch off encryption, an asynchronous background task is started which decrypts all encrypteddisk data.

When you use CREATE NEW KEY, the instance is forced to prepare a new random encryption key. This key willbegin to be used after the next savepoint has occurred. The time in which this takes to occur is depending onthe current workload of the SAP HANA database. No changes will be made to old written disk data.


136 2012-12-19

When you invoke APPLY CURRENT KEY, disk data that is encrypted with an old key is implicitly decryptedusing the old key and then encrypted again with the current key. You may wish to do this if an old key wascompromised or if you want to consolidate a set of old keys. When new key has been created with CREATENEW KEY and then APPLY CURRENT KEY is called, a savepoint is internally triggered so that the new key canbe used for encryption of old data.

Note:Currently only the finally written disk data will be encrypted. The redo log is not affected by thiscommand. For more information about the redo log please see the "Backing Up and Recovering theSAP HANA database" section in the "SAP HANA Administration guide" available from the SAP HANAAppliance page.

Example

You cause a non-encrypted instance to begin the asynchronous task of encrypting disk data.

ALTER SYSTEM PERSISTENCE ENCRYPTION ON

*

See Also

For more information about the use of Encryption in the SAP HANA database please see the "Managing DataVolume Encryption" section in the "SAP HANA Administration guide" available from the SAP HANA Appliancepage.

ALTER SYSTEM RECLAIM DATAVOLUME

Syntax

ALTER SYSTEM RECLAIM DATA VOLUME [SPACE] [<host_port>] <percentage_of_overload_size> <shrink_mode>

Syntax Elements

<host_port> ::= 'host_name:port_number'

Specifies the server on which the size of the persistency should be reduced.

<percentage_of_overload_size> ::= <int_const>

Specifies to which percentage of the overload size the data volume should be reduced.

<shrink_mode> ::= DEFRAGMENT | SPARSIFY

Specifies the strategy to reduce the persistency size. DEFRAGEMENT is the default. Please note thatSPARSIFY is not supported yet and reserved for future use.

Description

The command should be used when unused space inside the persistencies should be freed. It reduces datavolume size to a N% of overload size; it works like defragmenting a hard disk, pages scattered around thedata volume will be moved to the front of the data volume and the free space at the end of the data volumewill be truncated.

If <host_port> is omitted the statement is distributed to all servers with persistence.

Example

In the example below the persistencies of all servers in the landscape are defragmented and reduced to120% of the overload size.

ALTER SYSTEM RECLAIM DATAVOLUME 120 DEFRAGMENT

ALTER SYSTEM RECLAIM LOG

Syntax


137 2012-12-19




Description

You use this command when the database has accumulated a lot of log segments and you want to reclaimdisk space of currently unused log segments.

Log segment accumulation can be caused in several ways. For example, when the automatic log backup isnot operational for a long period or the log savepoint is blocked for an extended time. When such an issueoccurs you use the ALTER SYSTEM RECLAIM LOG command only after the root cause of the log accumulationhas been fixed.

Example

You reclaim disk space of currently unused log segments using the following command:


ALTER SYSTEM RECLAIM VERSION SPACE

Syntax


Description

This command triggers the row-store garbage collector to free up memory space and enhance systemresponsiveness.

You do not need to use this command if only short transactions are being executed on the system. This isbecause the garbage collector is automatically triggered when a transaction ends. If there are long-runningtransactions on the system, this command helps overall system performance if the m_mvcc_tablesmonitoring view shows a large number of row-store versions(e.g. over 1M).


m_mvcc_tables shows the number of row-store versions in the system.

Example

You check the number of row-store versions on the system using the m_mvcc_tables monitoring view.

SELECT * FROM m_mvcc_tables

You trigger row-store garbage collection.


ALTER SYSTEM RECONFIGURE SERVICE

Syntax

ALTER SYSTEM RECONFIGURE SERVICE (<service_name>,<host>,<port>)

Syntax Element

<service_name> ::= <string_literal>

The name of the service you wish to reconfigure. See M_SERVICE_TYPES monitoring view for a list ofavailable service types.

<host> ::= <string_literal> <port> ::= <unsigned_integer>

The host and port number where you would like to reconfigure a service.

Description


138 2012-12-19

You use ALTER SYSTEM RECONFIGURE SERVICE to reconfigures a specified service by applying the currentconfiguration parameters.

This command is used after changing multiple configuration parameters with the ALTER CONFIGURATIONcommand without the RECONFIGURE option set. See ALTER SYSTEM ALTER CONFIGURATION

To reconfigure a specific service specify <host> and <port> and leave <service_name> empty.

To reconfigure all services of a type, specify <service_name> and leave <host> and <port> empty.

To reconfigure all services, leave all parameters empty.

Example

You use the following command to reconfigure all services on the hana.hana.com host using port number30303:

ALTER SYSTEM RECONFIGURE SERVICE ('','hana.yourcompany.com',30303)

You use the following command to reconfigure all services of type indexserver:

ALTER SYSTEM RECONFIGURE SERVICE ('indexserver','',0)

See Also

ALTER SYSTEM ALTER CONFIGURATION

ALTER SYSTEM REMOVE TRACES

Syntax

ALTER SYSTEM REMOVE TRACES (<host>, <trace_type_name_list>)

Syntax Elements

<host> :== <string_literal>

The name of the host where the traces are to be removed.

<trace_type_name_list> ::= <trace_file> [{, trace_file>}...]

The trace file types to be removed.

<trace_file> :== !!see table below.

You set the <trace_file> parameter to one of the file specifications shown in the table below.

Trace Type <trace_file>

ALERT *a lert_*.trc

CLIENT loca lcl ient_*.trc

CRASHDUMP *.crashdump.*

EMERGENCYDUMP *.emergencydump.*

* a l l *.trc fi les of services l i s ted below

INDEXSERVER,NAMESERVER,...,DAEMON open *.trc fi les of a s ingle service type

Description

You use this command to Delete the trace files on a specified host to reduce disk space used by large tracefiles. When a service has a trace file open it cannot be deleted. In this case you can clear the trace file usingthe ALTER SYSTEM CLEAR TRACES command.

Example

You use the command below to delete all of the ALERT traces from the lu873.sap.com host.

ALTER SYSTEM REMOVE TRACES ('lu873.sap.com', '*alert_*.trc');


139 2012-12-19

See Also

ALTER SYSTEM CLEAR TRACES

ALTER SYSTEM RESET MONITORING VIEW

Syntax

ALTER SYSTEM RESET MONITORING VIEW <view_name>

Syntax Elements

<view_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The name of the resettable monitoring view to be reset.

Note:Not all monitoring views can be reset using this command. You can identify a view as being able tobe reset as its view name will have the suffix "_RESET".

Description

You use this command to reset statistics data for the specified monitoring view.

You use this command to define a starting point for your measurements. First, you reset the monitoring view,then execute an action. After the action is completed, query "_RESET" version of the monitor view to get thestatistical information gathered since the last reset.

Example

In the example below you reset the "SYS"."M_HEAP_MEMORY_RESET" monitoring view.

ALTER SYSTEM RESET MONITORING VIEW "SYS"."M_HEAP_MEMORY_RESET"

ALTER SYSTEM SAVE PERFTRACE

Note:Performance tracing is only to be used in conjunction with SAP Support personnel. The collectedperftrace data cannot be analyzed by an end user.

Syntax

ALTER SYSTEM SAVE PERFTRACE [INTO FILE <file_name>]

Syntax Elements

INTO FILE <file_name> <file_name> ::= <string_literal>

Specifies a file where raw performance data will be saved.

Description

You use this command to collect raw performance trace data from .prf files and save the information into asingle .tpt file. The .tpt file will be saved in the trace directory of your SAP HANA database instance. If you donot specify a file name then 'perftrace.tpt' will be used.

The performance trace data (.tpt) file can be downloaded from SAP HANA Studio. To obtain the trace files inthe Navigator pane you right click on the system you wish to diagnose. You select Administration from thecontext menu and in the dialog that appears you then select the Diagnosis Files tab. You right click on aTrace file and select Download.


The status of the performance trace can be monitored from M_PERFTRACE. While a trace file is saving it is shown as 'Save PerfTrace' job in M_JOB_PROGRESS.


140 2012-12-19

Cancellation

Saving a perftrace can take some time. You can cancel the job shown in M_JOB_PROGRESS with ALTERSYSTEM CANCEL [WORK IN] SESSION

Example

You save raw performance trace data into the 'mytrace.tpt' file.

ALTER SYSTEM SAVE PERFTRACE INTO FILE 'mytrace.tpt'

ALTER SYSTEM SAVEPOINT

Syntax


Description

Executes a savepoint on the persistence manager. A savepoint is a point in time when a complete consistentimage of the database is persisted to disk. The consistent image can be used to restart the database.

Normally a savepoint is executed periodically as configured by the configuration parametersavepoint_interval_s in section [persistence]. For special (normally test) purposes, the savepoint canbe disabled. In this case, you can use this command to manually execute a savepoint.


See M_SAVEPOINT_STATISTICS for savepoint statistics

Example

You execute a savepoint on the persistence manager.


ALTER SYSTEM START PERFTRACE


Syntax

ALTER SYSTEM START PERFTRACE [<user_name>] [<application_user_name>] [<application_name>] [PASSPORT_TRACELEVEL <passport_level>] [PLAN_EXECUTION] [FUNCTION_PROFILER] [DURATION <duration_seconds>]

Syntax Element

<user_name> ::= USER <identifier>

Restricts perftrace collection to an SQL user name

<application_user_name> ::= APPLICATIONUSER <identifier>

Restricts perftrace collection to the application user name.

<application_name> ::= APPLICATION <identifier>

Restricts perftrace collection to the application name.

<passport_level> ::= MEDIUM | HIGH

Specifies the level of trace data to be collected. This is effectively a filter controlling the amount of datacollected in end-to-end scenarios.


141 2012-12-19

PLAN_EXECUTION

Specifies that plan execution details should be traced.

FUNCTION_PROFILER

Specifies that function level details should be traced.

<duration_seconds> ::= <numeric literal>

The number of seconds which perftrace should be collected. After this period expires perftrace isautomatically stopped. If you do not specify this parameter you must stop perftrace with the ALTER SYSTEMSTOP PERFTRACE statement.

Description

The ALTER SYSTEM START PERFTRACE command starts performance tracing.

Only one perftrace can be active at a time.


The status of performance trace can be monitored from M_PERFTRACE.

Example

You start performance tracing for the user sql_user on application user name app_user for applicationapp_name. You also specify that plan execution and function level detail should be traced.

ALTER SYSTEM START PERFTRACE USER sql_user APPLICATIONUSER app_user APPLICATION app_name PLAN_EXECUTION FUNCTION_PROFILER

ALTER SYSTEM STOP PERFTRACE


Syntax


Description

Stops an active performance trace. After stopping the trace, you collect and save the performance tracedata with ALTER SYSTEM SAVE PERFTRACE.

Example

You stop an active performance trace.


ALTER SYSTEM STOP SERVICE

Syntax

ALTER SYSTEM STOP SERVICE <host_port> [IMMEDIATE [WITH COREFILE]]

Syntax Element

<host_port> ::= <host_name:port_number> | ('<host_name>',<port_number>)

Host and port of the service to be stopped.

IMMEDIATE


142 2012-12-19

Immediately stop the service without waiting for regular shutdown.

WITH COREFILE

Write a core dump file.

Description

Stop single or multiple services on the designated host.

Note:Typically a service you stop will be automatically restarted by the SAP HANA database system.

You should use this command after changing a configuration parameter that can not be changed whilst theSAP HANA database is running.

Example

You stop a service running on host hdb1.yourcompany.com on port 30303.

ALTER SYSTEM STOP SERVICE 'hdb1.yourcompany.com:30303'

SET SYSTEM LICENSE

Syntax

SET SYSTEM LICENSE <license key>

Syntax Elements

<license key> ::= <string_literal>

The license key to be set.

Description

Installs a license key to the database instance. The license key (<license key>="">) is to be copied and pastedfrom the license key file. The system privilege LICENSE ADMIN is required to execute this command.

Example

Note:The license shown below does not contain a license key, so you cannot run this example.

You set the system license with a license code.

SET SYSTEM LICENSE '----- Begin SAP License ----- SAPSYSTEM=HD1 HARDWARE-KEY=K4150485960 INSTNO=0110008649 BEGIN=20110809 EXPIRATION=20151231 LKEY=... SWPRODUCTNAME=SAP-HANA SWPRODUCTLIMIT=2147483647 SYSTEM-NR=00000000031047460'

UNSET SYSTEM LICENSE ALL

Warning:Running this command immediately locks down your SAP HANA database. You will require a newlicense key before you can use the system further.

Syntax


Description


143 2012-12-19

Deletes all currently installed license keys. After using this command the SAP HANA database will be lockeddown immediately and will require a new valid license key before it can be used further. The system privilegeLICENSE ADMIN is required to execute this command.

Example

You delete all currently installed license keys.


Session Management Statements

CONNECT

Syntax

CONNECT <connect_option>

Syntax Elements

<connect_option> ::= <user_name> PASSWORD <password> | WITH SAML ASSERTION '<xml>' <user_name> ::= <simple_identifier>

<password> ::= <letter_or_digit>...

Description

Connect to the database instance by specifying user_name and password or by specifying a SAML assertion.

Example

You create a SAML provider with the name ac_saml_provider specifying a subject and issuer for ACompany.

CREATE SAML PROVIDER ac_saml_provider WITH SUBJECT 'CN = wiki.detroit.ACompany.corp,OU = ACNet,O = ACompany,C = EN' ISSUER 'E = [email protected],CN = ACNetCA,OU = ACNet,O = ACompany,C = EN';

You create a user called new_user with password Password1. The user can connect to the system using thegiven password and with an assertion of the SAML provider ac_saml_provider.

CREATE USER new_user PASSWORD Password1 WITH IDENTITY ANY FOR SAML PROVIDER ac_saml_provider;

You connect to the database with the user name new_user and password Password1.

CONNECT new_user PASSWORD Password1

SET HISTORY SESSION

Syntax

SET HISTORY SESSION TO <when>

Elements

<when>:

User should specify an exact moment when the session is to travel

<when> ::= NOW | COMMIT ID <commit_id> | UTCTIMESTAMP <timestamp> <commit_id> ::= <unsigned_integer>

Description

SET HISTORY SESSSION makes current session to see a previous version of history tables. User may specify theversion in COMMIT ID or UTCTIMESTAMP format, or may get back to current version by specify NOW. After


144 2012-12-19

SET HISTORY SESSION command with a COMMIT ID or UTCTIMESTAMP is issued, current session sees an oldversion of history tables. If NOW option is given, current session gets back to a normal session, and seescurrent version of history tables. This command only works for history tables, and visibility on normal tables isnot affected.


The last commit id can be obtained from m_history_index_last_commit_id after each commit.The timestamp of the last commit can be obtained from m_transaction_history.

Example

You create a history column table x, so that you can time travel through the table's history.

create history column table x ( a int, b int ); // after turnning off auto commit

You insert some values into table x

insert into x values (1,1); commit; insert into x values (2,2); commit;

You obtain the commit id of the last commit.

select last_commit_id from m_history_index_last_commit_id where session_id = current_connection;

Note:please replace <last_comit_id> in the steps below with the value you obtained in this step.

You delete all the data from table x.

delete from x; commit;

You time travel into the history of table x before you deleted the data and perform a select.

set history session to commit id <last_comit_id>; select * from x;

The select query above returns the two records (1,1) and (2,2), which existed before you deleted thecontents of table x.

You obtain the timestamp of <last_comit_id>.

select commit_time from sys.transaction_history where commit_id = <last_comit_id>;

Note:please replace <commit_time> in the steps below with the value you obtained in this step.

You time travel into the history of table x using a timestamp and perform a select.

set history session to utctimestamp '<commit_time>'; select * from x;

The select query above returns the two records (1,1) and (2,2), which existed before you deleted thecontents of table x.

You set the history session to the current commit state and perform a select.

set history session to now; select * from x;

The select query returns an empty table because, at the current state, the table has had all of its datadeleted.

SET SCHEMA

Syntax


145 2012-12-19

SET SCHEMA <schema_name>

Syntax Elements

<schema_name> ::= <string_literal>

The name of the schema that the session should change to.

Description

Changes the current schema of the session.

The current schema is used when SQL statements use database object names, for example table names, thatare not prefixed with a schema name.

Example

You create new schema called MY_SCHEMA.

CREATE SCHEMA MY_SCHEMA;

You change the current schema of the session to MY_SCHEMA.

SET SCHEMA MY_SCHEMA;

SET [SESSION]

Syntax

SET [SESSION] <key> = <value>

Syntax Elements



<value> ::= <string_literal>

The desired value of a session variable. The maximum length of value is 512 characters.

Description

With this command you can set session variables of your own database session by providing key and valuepairs.

Note:There are several available read-only session variables that you cannot change with this command:APPLICATION, APPLICATIONUSER, TRACEPROFILE.

Session variables can be retrieved using SESSION_CONTEXT function and unset using the UNSET [SESSION]command.

Example

You set the session variable MY_VAR to 'abc'.


You select the variable MY_VAR from the current session.


You unset the session variable MY_VAR.

UNSET 'MY_VAR';


146 2012-12-19

UNSET [SESSION]

Syntax

UNSET [SESSION] <key>

Syntax Elements



Description

Using UNSET [SESSION] you can unset session variables of the current session.

Note:There are several read-only session variables and they are APPLICATION, APPLICATIONUSER,TRACEPROFILE.

Example

You set the session variable MY_VAR to 'abc'.


You select the variable MY_VAR from the current session.


You unset the session variable MY_VAR.

UNSET 'MY_VAR';

Transaction Management Statements

COMMIT

Syntax

COMMIT

Description

The system supports transactional consistency which guarantees current job to be either completely appliedto the system or disposed. If a user wants to apply current job to the system persistently, the user shouldissue COMMIT command. If COMMIT command is issued and successfully processed, any change on thesystem which current transaction has done is applied to the system and the change will be visible to otherjobs which will start in the future. The job which has already committed via COMMIT command cannot bereverted. In a distributed system, standard 2-phase-commit protocol is complied. In the first phase,transaction coordinator consults every participant whether if it is ready to commit, and sends the result tothe participants in the second phase. COMMIT command only works with 'autocommit' disabled session.

Example

COMMIT

LOCK TABLE

Syntax

LOCK TABLE <table_name> IN EXCLUSIVE MODE [NOWAIT]

Syntax Elements

<table_name> ::= [<schema_name>.]<identifier>


147 2012-12-19

<schema_name> ::= <identifier>

For descriptions on schema name and identifier, please refer to Identifiers.

Description

LOCK TABLE command acquires an exclusive lock for a table.

If the NOWAIT option is specified and LOCK TABLE fails to acquire lock an error code is returned.

Note:When the error is generated the current transaction is not rolled-back.

LOCK MODE: EXCLUSIVE

Can be acquired by LOCK TABLE command explicitly or DDL command implicitly The transaction that holds the lock can read and write the table. Only the transaction that holds the lock can access the table. Lock requests for the table by other transactions are blocked while the EXCLUSIVE lock is held. The database releases acquired locks at the end of the transaction.

LOCK MODE: INTENTIONAL EXCLUSIVE

Acquired by DML implicitly. Multiple transactions can acquired a INTENTIONAL EXCLUSIVE lock. EXCLUSIVE lock requests for the table by other transactions are blocked while the INTENTIONALEXCLUSIVE lock is held. The database releases acquired locks at the end of the transaction.


The status of lock acquisition can be monitored with M_TABLE_LOCKS.

Example

You create table A.


You acquire an exclusive lock for table A.

LOCK TABLE A IN EXCLUSIVE MODE

You attempt to acquire an exclusive lock for table A and specify that an error should be returned if the lockcannot be immediately obtained.

LOCK TABLE A IN EXCLUSIVE MODE NOWAIT

ROLLBACK

Syntax

ROLLBACK

Description

The SAP HANA database supports transactional consistency which guarantees that a transaction will becompletely applied to the system or disposed. During a transaction data manipulation language (DML)modifications to the database can be explicitly reverted via ROLLBACK command. After ROLLBACK commandis issued, changes made during the current transaction are reverted and current database session is set to anidle state. ROLLBACK command only works with an autocommit disabled session.

If you attempt to use the ROLLBACK statements in an autocommit enabled session, nothing will occur astransactions are automatically committed to the database.

Example

Note:Before attempting to execute the example below, please ensure that you are using an autocommitdisabled session.


148 2012-12-19

You create a table T.

CREATE TABLE T (KEY INT PRIMARY KEY, VAL INT);

You insert three rows into table T.

INSERT INTO T VALUES (1, 1); INSERT INTO T VALUES (2, 2); INSERT INTO T VALUES (3, 3);

You rollback the current transaction.

ROLLBACK;

You select the data in table T.

SELECT * FROM T;

The SELECT command above returns an empty table. From this you can see that the data definition language(DDL) command used to create the table has persisted, but the DML used to create the table data has beenrolled back.

See Also

For more information on changing the autocommit status of your database session, please see "Customizingthe Administration Console" section in the "SAP HANA Administration guide" available from the SAP HANAAppliance page.

SET TRANSACTION

Syntax

SET TRANSACTION <isolation_level> | <transaction_access_mode>

Syntax Elements

isolation_level ::= ISOLATION LEVEL <level>

The isolation level sets the statement level read consistency of the data in the database. If <isolation_level> isommited the default is READ COMMITTED

level ::= READ COMMITTED | REPEATABLE READ | SERIALIZABLE

READ COMMITTED The READ COMMITTED isolation level provides statement level read consistency during a transaction. Eachstatement in a transaction sees commited state of the data in the database as the execution of thestatement begins. This means that each statement in the same transaction may see varying snapshots of thedata in the database as they are executed as data can be commited during the transaction.

REPEATABLE READ/SERIALIZABLE The REPEATABLE READ/SERIALIZABLE isolation level provides transaction level snapshot isolation. Allstatements of a transaction see the same snapshot of the database data. This snapshot contains all changesthat were committed at the time the transaction started along with the changes made by the transactionitself.

transaction_access_mode ::= READ ONLY | READ WRITE

The SQL-transaction access mode controls if a transaction can modify data during execution. Iftransaction_access_mode is ommited the default is READ ONLY.

READ ONLY When READ ONLY access mode is set, then only read operations with SELECT statements are allowed. Anexception will be thrown if update or insert operations are attempted whilst in this mode.

READ WRITE When READ WRITE access mode is set, statements within a transaction can freely read or make changes tothe database data as required.

Description


149 2012-12-19


Description

The SAP HANA database uses multi-version concurrency control (MVCC) to ensure consistent read operations.Concurrent read operations see a consistent view of the database data without blocking concurrent writeoperations. Updates are implemented by inserting new versions of data and not by overwriting existingrecords.

The isolation level specified determines the lock operation type that will be used. The system supports bothstatement level snapshot isolation and transaction level snapshot isolation.

For statement snapshot isolation use level READ COMMITTED.For transaction snapshot isolation use REPEATABLE READ or SERIALIZABLE.

During a transaction when rows are inserted, updated or deleted, the system sets exclusive locks on theaffected rows for the duration of the transaction. The system also sets shared locks on the affected tables forthe duration of the transaction. This guarantees that the table is not dropped or altered while rows of thetable are being updated. The database releases these locks at the end of the transaction.

Note:Reading a row does not set any locks on either tables or rows within the database regardless of theisolation level used.

Data Definition Language and Transaction Isolation

Data Definition Language (DDL) statements (CREATE TABLE, DROP TABLE, CREATE VIEW, etc) always takean instantaneous effect on following SQL statements regardless of the transaction isolation level being used.For an example of this behaviour please consider the following sequence:

1. A long running SERIALIZABLE isolation transaction begins operating on Table C.2. From outside the transaction some DDL is executed which adds a new column to Table C.3. From within the SERIALIZABLE isolation transaction the new column will be accessible as soon as the

DDL statement completes. This access occurs regardless of the isolation level of the transaction.

Example

You set the transaction isolation level to READ COMMITTED to provide statement level read consistencyduring the current transaction.

SET TRANSACTION ISOLATION LEVEL READ COMMITTED;

Access Control Statements

ALTER SAML PROVIDER

Syntax

ALTER SAML PROVIDER <saml_provider_name> SET SUBJECT <subject_distinguished_name> ISSUER <issuer_distinguished_name>

Syntax Elements

<saml_provider_name> ::= <simple_identifier>

The identifier of a SAML provider to be altered.

<subject_distinguished_name> ::= <string_literal>

The subject name provided in the certificate of the SAML identity provider.

<issuer_distinguished_name> ::= <string_literal>

The issuer name provided in the certificate of the SAML identity provider.

Description

The ALTER SAML PROVIDER statement changes the property of a SAML provider known to the SAP HANAdatabase. <saml_provider_name> has to be an existing SAML provider created using the CREATE SAMLPROVIDER command.


150 2012-12-19

Only database users having the system privilege USER ADMIN are allowed to change a SAML provider.


SAML_PROVIDERS: shows all SAML providers with their subject name and issuer_name.

Example

You create the ac_saml_provider SAML provider.


You view the current defined SAML providers in the SAP HANA database.

select * from sys.SAML_PROVIDERS

You alter the ac_saml_provider.

ALTER SAML PROVIDER ac_saml_provider SET SUBJECT 'CN = wiki.detroit.BCompany.corp,OU = BCNet,O = BCompany,C = EN' ISSUER 'E = [email protected],CN = BCNetCA,OU = BCNet,O = BCompany,C = EN';

You check the current defined SAML providers in the SAP HANA database. You should now see the updatedSAML provider.

select * from sys.SAML_PROVIDERS

See Also

CREATE SAML PROVIDER

ALTER USER

Syntax

ALTER USER <user_name> <alter_user_option>

Syntax Elements

<user_name> ::= <simple_identifier>

<alter_user_option> ::= PASSWORD <password> [<user_parameter_option>] | <user_parameter_option> | IDENTIFIED EXTERNALLY AS <external_identity> [<user_parameter_option>] | RESET CONNECT ATTEMPTS | DROP CONNECT ATTEMPTS | DISABLE PASSWORD LIFETIME | FORCE PASSWORD CHANGE | DEACTIVATE [USER NOW] | ACTIVATE [USER NOW] | DISABLE <authentication_mechanism> | ENABLE <authentication_mechanism> | ADD IDENTITY <provider_identity>... | ADD IDENTITY <external_identity> FOR KERBEROS | DROP IDENTITY <provider_info>... | DROP IDENTITY FOR KERBEROS


<user_parameter_option> ::= <set_user_parameters> [<clear_user_parameter_option>] | <clear_user_parameter_option>

<set_user_parameters> ::= SET PARAMETER CLIENT = <string_literal>

<clear_user_parameter_option> ::= CLEAR PARAMETER CLIENT | CLEAR ALL PARAMETERS

<external_identity> ::= <simple_identifier> | <string_literal>


151 2012-12-19

<authentication_mechanism> ::= PASSWORD | KERBEROS | SAML

<provider_identity> ::= <mapped_user_name> FOR SAML PROVIDER <saml_provider_name> | <external_identity> FOR KERBEROS <mapped_user_name> ::= ANY | <string_literal> <saml_provider_name> ::= <simple_identifier>

<provider_info> ::= FOR SAML PROVIDER <saml_provider_name>

Description

The ALTER USER statement modifies the database user. <user_name> must specify an existing databaseuser. Each user can do the ALTER USER statement for his own. But not all of the <alter_user_option>s can bespecified by the user himself. The ALTER USER statement for other users with all <alter_user_option>s canonly be done by users having the system privilege USER ADMIN.

Users created with PASSWORD cannot be changed to EXTERNALLY and vice versa: the users created withEXTERNALLY cannot be changed to PASSWORD. But their <password> or their <external_identity> can bechanged.

You can change a user's password with this command. Password must follow the rules defined for the currentdatabase. The password rules include the minimal password length and the definition which of the charactertypes ( lower, upper, digit, special characters ) have to be part of the password. The password for a user hasto be changed regularly according to the password policy specified for the database instance or changed bythe user himself when first connecting to the database instance.

You can change the external authentication. External users are authenticated using an external system, e.g.a Kerberos system. Such users do not have a password, but , e.g. a Kerberos principal name. For detailedinformation about external identities, contact your domain administrator.

<user_parameter_option> can be used to set, change or clean the user parameter CLIENT.<set_user_parameters> can be used to set the user parameter CLIENT for a user in the database.When using reports this user parameter CLIENT can be used to restrict the access rights of user <user_name>to info concerning the specified client. The <user_parameter_option> can not be specified by the user himself.

If the number of MAXIMUM_INVALID_CONNECT_ATTEMPTS (see monitoring view M_PASSWORD_POLICY) isreached before a successful (correct user/password-combination) connect is done, then this user is locked forsome minutes before being allowed to connect (even with correct user/password-combination) again. Withthe command ALTER USER <user_name> RESET CONNECT ATTEMPTS a user with system privilege USER ADMIN can reset the number of invalid attempts to 0 and therefore allowthe user to connect immediately.

Information on invalid connect attempts having happened can be checked in system viewINVALID_CONNECT_ATTEMPTS. With the commandALTER USER <user_name> DROP CONNECT ATTEMPTS a user with system privilege USER ADMIN or the user himself can delete the information of invalid connectattempts having happened.

With the commandALTER USER <user_name> DISABLE PASSWORD LIFETIME a user with system privilege USER ADMIN can exclude user <user_name> from all password-life-time-checks.This should be used only for technical users, not for normal database users. Please check the description ofconfiguration parameters concerning the password policy.

With the commandALTER USER <user_name> FORCE PASSWORD CHANGE a user with system privilege USER ADMIN can force user <user_name> to change his password immediatelyafter the next connect before being allowed to work any further.

With the commandALTER USER <user_name> DEACTIVATE USER NOW a user with system privilege USER ADMIN can deactivate / lock the user account of user <user_name>. Afterthe user account is deactivated / locked, the user cannot connect to the SAP HANA database. To re-activate /unlock user <user_name> a user with system privilege USER ADMIN has either to use the command ALTER


152 2012-12-19

USER <user_name> ACTIVATE USER NOW or, in case of a user with PASSWORD authentication mechanism,to reset the user's password with ALTER USER <user_name> PASSWORD <password>. With the commandALTER USER <user_name> ACTIVATE USER NOW a user with system privilege USER ADMIN can re-activate / unlock the user account of user <user_name>who had been deactivated before.


Conguration parameters concerning the password can be checked with the monitoring viewM_PASSWORD_POLICY. These parameter are stored in indexserver.ini, section 'password policy'. The description of the parameters concerned can be found in SAP HANA Security Guide, Appendix, PasswordPolicy Parameters.


USERS: shows all users, their creator, creation date and some info about their current states. USER_PARAMETERS: shows the defined user_parameters; currently only CLIENT is available. INVALID_CONNECT_ATTEMPTS: shows how many invalid connect attempts were made for each user. LAST_USED_PASSWORDS: shows info about dates of last password-changes per user. M_PASSWORD_POLICY: shows configuration parameters describing the allowed layout of the password andits lifetimes.

Examples

You create a SAML provider named ac_saml_provider in the database specifying a subject and issuer forACompany.


You create user named new_user that can connect using a password or with an assertion of the SAMLprovider ac_saml_provider. The <mapped_user_name> was set to ANY as the assertion will provide thedatabase user name.


You force the user to change their password.

ALTER USER new_user FORCE PASSWORD CHANGE;

You disable SAML authentication for the user.

ALTER USER new_user DISABLE SAML;

You reset the number of invalid connect attempts to zero for new_user.

ALTER USER new_user RESET CONNECT ATTEMPTS;

You define the external identity for KERBEROS and enable KERBEROS for this user.

ALTER USER new_user ADD IDENTITY 'testkerberosName' FOR KERBEROS; ALTER USER new_user ENABLE KERBEROS;

Note:Adding an external identification mechanism to a user does not automatically enable it. You must dothis as a separate step as shown in this example.

You remove ac_saml_provider SAML provider identity from new_user.

ALTER USER new_user DROP IDENTITY FOR SAML PROVIDER ac_saml_provider;

You disable the account.

ALTER USER new_user DEACTIVATE;

CREATE ROLE


153 2012-12-19

Syntax

CREATE ROLE <role_name>

Syntax Elements

<role_name> ::= <identifier>

Description

The CREATE ROLE statement creates a new role. Only database users with the system privilege ROLE ADMIN are allowed to create roles. The specified role name must not be identical to the name of an existing user or role.

A role is a named collection of privileges and can be granted to either a user or a role. If you want to allowseveral database users to perform the same actions, you can create a role, grant the needed privileges tothis role, and then grant the role to the database users. Every user is allowed to grant privileges to an existing role, but only users having system privilege ROLEADMIN are allowed to grant roles to roles and users.

There are 5 roles that are delivered with the SAP HANA database:

PUBLIC Every database user is granted this role automatically. This role contains read-only access to system views, monitoring views and execute privileges for someprocedures. These privileges cannot be revoked. This role can be granted further privileges which can then be revoked afterwards.

MODELING This role contains all privileges required for using the information modeler in the SAP HANA studio.

CONTENT_ADMIN This role contains the same privileges as the MODELING role, but with the addition that this role is allowed togrant these privileges to other users. In addition, it contains the repository privileges to work with importedobjects.

MONITORING This role contains privileges for full read-only access to all meta data, the current system status in system andmonitoring views, and the data of the statistics server.

SUPPORT This role is to be used for SAP support.This role contains privileges for full read-only access to all meta data, the current system status in system andmonitoring views, and the data of the statistics server. Additionally it contains the privileges to access thebase information of the system and monitoring views. This information is usually only accessible by theSYSTEM user.To ensure this role is used only for support there are some restrictions place on it:

it cannot be granted to database user SYSTEMit cannot be granted to more than one database user at a timeit cannot be granted to another roleno role can be granted to itonly systemprivileges (see GRANT statement) can be granted to this role. However for securityreasons SAP recommends systemprivileges should be granted to the user and not to the SUPPORTrole.with every upgrade of the SAP HANA database the privileges in this role are reset the default settings.


ROLES: displays all roles, their creator and creation date. GRANTED_ROLES: displays the roles that are granted to users or roles. GRANTED_PRIVILEGES: displays the privileges that are granted to users or roles.

Example

You create a role with the name role_for_work_on_my_schema.


154 2012-12-19

CREATE ROLE role_for_work_on_my_schema;

CREATE SAML PROVIDER

Syntax

CREATE SAML PROVIDER <saml_provider_name> WITH SUBJECT <subject_distinguished_name> ISSUER <issuer_distinguished_name>

Syntax Elements


The identifier of a SAML provider to be created.

<subject_distinguished_name> ::= <string_literal>

The subject name provided in the certificate of the SAML identity provider.

<issuer_distinguished_name> ::= <string_literal>

The issuer name provided in the certificate of the SAML identity provider

Description

The CREATE SAML PROVIDER statement defines a SAML provider that will be known to the SAP HANAdatabase. <saml_provider_name> has to be unique from any existing SAML provider. Only database users having the system privilege USER ADMIN are allowed to create a SAML provider. Eachuser having this privilege is also allowed to drop any SAML provider.

A SAML provider is required to be able to provide a SAML connection facility for users.

The <subject_distinguished_name> and the <issuer_distinguished_name> are the X.500 distinguished namescorresponding to the subject and issuer of the X.509 certificate used by the SAML identity provider to signassertions. The syntax of these names can be found in ISO/IEC 9594-1. A detailed description of the concepts of SAML can be found in Oasis SAML 2.0.


SAML_PROVIDERS: shows all SAML providers with their subject name and issuer name.

Example

You create a SAML provider with the name ac_saml_provider in the database specifying the subject andissuer to belong to ACompany.


CREATE USER

Syntax

CREATE USER <user_name> [PASSWORD <password>] [IDENTIFIED EXTERNALLY AS <external_identity>] [WITH IDENTITY <provider_identity>...] [<set_user_parameters>]

Syntax Elements



<external_identity> ::= <simple_identifier> | <string_literal>

<provider_identity> ::= <mapped_user_name> FOR SAML PROVIDER <saml_provider_name>


155 2012-12-19

| <external_identity> FOR KERBEROS

<mapped_user_name> ::= ANY | <string_literal>


<set_user_parameters> ::= SET PARAMETER CLIENT = <string_literal>

Description

The CREATE USER statement creates a new database user. Only database users having the system privilege USER ADMIN are allowed to create another database user. The specified user name must not be identical to the name of an existing user, role, or schema. There are some users, which are delivered with the SAP HANA database: SYS, SYSTEM, _SYS_REPO,_SYS_STATISTICS.

Users in the database can be authenticated by different mechanism, the internal authentication mechanismusing password and by external mechanism, like Kerberos or SAML. One user can be authenticated by morethan one mechanism at one time, but only one password and one external_identity can be valid at one time.In opposite to this, more than one <provider_identity> can exist for one user at one time. At least oneauthentication mechanism has to be specified to allow the user to connect and work with the databaseinstance.For compatibility reasons the syntax IDENTIFIED EXTERNALLY AS <external_identity> will stay alive and dothe same as the syntax <external_identity> FOR KERBEROS.

Password must follow the rules defined for the current database. The password rules include the minimalpassword length and the definition which of the character types ( lower, upper, digit, special characters )have to be part of the password. The password for a user has to be changed regularly according to thepassword policy specified for the database instance.The password provided during CREATE USER command will be taken as provided, whereas the <user_name>will be changed to upper_case as every <simple_identifier>.

External users are authenticated using an external system, e.g. a Kerberos system. Such users do not have apassword, but , e.g. a Kerberos principal name. For detailed information about external identities, contactyour domain administrator.

If ANY is given as mapped user name, the SAML assertion will contain the name of the database user, theassertion is valid for. <saml_provider_name> must specify an existing SAML provider.

<set_user_parameters> can be used to set the user parameter CLIENT when a user is created in thedatabase. When using reports this user parameter CLIENT can be used to restrict the access rights of user<user_name> to info concerning the specified client.

For each database user a schema with the user's name is created implicitly. This can not be droppedexplicitly. It will be dropped when the user is dropped. The database user owns that schema and uses it as hisdefault schema, when he does not specify a schema name explicitly.


Conguration parameters concerning the password can be checked with the monitoring viewM_PASSWORD_POLICY. These parameter are stored in indexserver.ini, section 'password policy'. The description of the parameters concerned can be found in SAP HANA Security Guide, Appendix, PasswordPolicy Parameters.


USERS: shows all users, their creator, creation date and some info about their current states. USER_PARAMETERS: shows the defined user_parameters; currently only CLIENT is available. INVALID_CONNECT_ATTEMPTS: shows how many invalid connect attempts were made for each user. LAST_USED_PASSWORDS: shows info about dates of last password-changes per user. M_PASSWORD_POLICY: shows configuration parameters describing the allowed layout of the password andits lifetimes. SAML_PROVIDERS: shows the existing SAML provider SAML_USER_MAPPING: shows the user's mappingnames for each SAML provider

Examples

Example 1 - Create user with password.

You create a user T12345 with a password Password123.


156 2012-12-19

You create a user T12345 with a password Password123.

CREATE USER T12345 PASSWORD Password123

Example 2 - Create user that uses an external authentication mechanism.

You create a SAML provider named ac_saml_provider in the database, specifying a subject and issuer forACompany.


You create a user called new_user with password Password1. The user can connect to the system using thegiven password and with an assertion of the existing SAML provider ac_saml_provider. The<mapped_user_name> is set to ANY as the assertion will provide the database user name.


DROP ROLE

Syntax

DROP ROLE <role_name>

Syntax Elements


Description

The DROP ROLE statement drops a role. <role_name> must specify an existing role. Only database users having the system privilege ROLE ADMIN are allowed to drop a role. Each user havingthis privilege is allowed to drop any role. Only those roles which are delivered with the SAP HANA databasecannot be dropped: PUBLIC, CONTENT_ADMIN, MODELING, MONITORING and SUPPORT.

If a role was granted to a user or role, it is revoked when the role is dropped. Revoking a role may lead tomaking some views inaccessible or making procedures not executable. This will occur if a view or proceduresdepends on any privilege of the dropped role.


ROLES: displays all roles, their creator and creation date. GRANTED_ROLES: displays the roles that are granted to users or roles. GRANTED_PRIVILEGES: displays the privileges that are granted to users or roles.

Example

A role with the name role_for_work_on_my_schema is created and dropped immediately afterwards;


DROP ROLE role_for_work_on_my_schema;

DROP SAML PROVIDER

Syntax

DROP SAML PROVIDER <saml_provider_name>

Syntax Elements


Description


157 2012-12-19

The DROP SAML PROVIDER statement drops the specified SAML provider. <saml_provider_name> has to bean existing SAML provider. If the SAML provider specified is used by a SAP HANA database user, the SAMLprovider cannot be dropped. Only database users having the system privilege USER ADMIN are allowed to drop a SAML provider.


SAML_PROVIDERS: shows all SAML providers with their subject name and issuer_name.

Example

You create SAML provider named ac_saml_provider, specifying a subject and issuer for ACompany.


You drop the SAML provider ac_saml_provider.

DROP SAML PROVIDER ac_saml_provider

DROP USER

Syntax

DROP USER <user_name> [<drop_option>]

Syntax Elements



Default = RESTRICT

Description

The DROP USER statement deletes a database user. <user_name> must specify an existing database user.

Only database users having the system privilege USER ADMIN are allowed to drop a database user. Each userhaving this privilege is allowed to drop any user. Only those users which are delivered with the SAP HANAdatabase can not be dropped: SYS, SYSTEM, _SYS_REPO, _SYS_STATISTICS.

If the <drop_option> RESTRICT is specified implicitly or explicitly then the user will not be dropped in case heis the owner of any other object than the schema with his name and other schemas created by him or in casethere is an object stored in one of his schemas which was not created by him. If the <drop_option> CASCADE is specified, the schema with the user's name and the schemas belonging tothe user, together with all objects stored in them (even if they are created by other users), are deleted.Objects owned by the user, even if they are part of another schema, are deleted. Objects that are dependenton deleted objects are deleted. Even public synonyms owned by the deleted user are deleted.Privileges on deleted objects are revoked. Privileges granted by the deleted user are revoked. Revokeprivileges may cause further revokes if they had been granted further.

Users created by the deleted user and roles created by him are not deleted. Audit policies created by the deleted user are not deleted.

It is possible to delete a user even if an open session of this user exists.


The deleted user will be deleted in any of these views:USERS: shows all users, their creator, creation date and some info about their current states. USER_PARAMETERS: shows the defined user_parameters; currently only CLIENT is available. INVALID_CONNECT_ATTEMPTS: shows how many invalid connect attempts were made for each user. LAST_USED_PASSWORDS: shows info about dates of last password-changes per user.

* The deletion of objects may influence all of the system views describing objects, like TABLES, VIEWS,PROCEDURES, ... .


158 2012-12-19

The deletion of objects may influence the view describing privileges like GRANTED_PRIVILEGES and all of themonitoring views like M_RS_TABLES, M_TABLE_LOCATIONS, ...

Example

A user with user_name NEW_USER was created before with, for example, this statemant:

CREATE USER new_user PASSWORD Password1;

The existing user new_user will now be dropped, together with all of his objects;

DROP USER new_user CASCADE;

GRANT

Syntax

GRANT <system_privilege>,... TO <grantee> [WITH ADMIN OPTION] | GRANT <schema_privilege>,... ON SCHEMA <schema_name> TO <grantee> [WITH GRANT OPTION] | GRANT <object_privilege>,... ON <object_name> TO <grantee> [WITH GRANT OPTION] | GRANT <role_name>,... TO <grantee> [WITH ADMIN OPTION] | GRANT STRUCTURED PRIVILEGE <privilege_name> TO <grantee>

Syntax Elements

<system_privilege> ::= AUDIT ADMIN | BACKUP ADMIN | CATALOG READ | CREATE SCENARIO | CREATE SCHEMA | CREATE STRUCTURED PRIVILEGE | DATA ADMIN | EXPORT | IMPORT | INIFILE ADMIN | LICENSE ADMIN | LOG ADMIN | MONITOR ADMIN | OPTIMIZER ADMIN | RESOURCE ADMIN | ROLE ADMIN | SAVEPOINT ADMIN | SCENARIO ADMIN | SERVICE ADMIN | SESSION ADMIN | STRUCTUREDPRIVILEGE ADMIN | TRACE ADMIN | TRUST ADMIN | USER ADMIN | VERSION ADMIN | <identifier>.<identifier>

System privileges are used to restrict administrative tasks. The table below describes the supported systemprivileges.


159 2012-12-19

System Privilege Privilege Purpose Comment

AUDIT ADMIN

Controls the execution of thefol lowing audi ting-relatedcommands : CREATE AUDIT POLICY,DROP AUDIT POLICY and ALTER AUDITPOLICY and the changes of theaudi ting configuration.

BACKUP ADMINAuthorizes the BACKUP command todefine and ini tia te a backup processor to perform a recovery process .

CATALOG READAuthorizes users to have unfi l teredread-only access to a l l sys tem andmonitoring views .

Normal ly, the content of these views i sfi l tered based on the privi leges of theaccess ing user.

CREATE SCENARIOControls the creation of ca lculationscenarios and cubes (ca lculationdatabase)

CREATE SCHEMA

Authorizes the creation of databaseschemas us ing the CREATE SCHEMAcommand.By defaul t each user ownsone schema, wi th this privi lege theuser i s a l lowed to create addi tiona lschemas .

CREATE STRUCTUREDPRIVILEGE

Authorizes the creation of StructuredPrivi leges (Ana lytica l Privi leges )

Only the owner of an Ana lytica l Privi legecan further grant or revoke that privi legeto other users or roles .

DATA ADMIN

Authorizes reading a l l data in thesys tem and monitoring views . It a l soenables execution of any DataDefini tion Language (DDL)commands in the SAP HANA database

A user having this privi lege cannot selector change data s tored tables for whichthey do not have access privi leges , butthey can drop tables or modi fy tabledefini tions .

EXPORT

Authorizes export activi ty in thedatabase via the EXPORT TABLEcommand.Note that bes ide thisprivi lege the user requires the SELECTprivi lege on the source tables to beexported.

IMPORTAuthorizes the import activi ty in thedatabase us ing the IMPORTcommands .

Note that bes ide this privi lege the userrequires the INSERT privi lege on the targettables to be imported.

INIFILE ADMINAuthorizes changing of sys temsettings .

LICENSE ADMINAuthorizes the SET SYSTEM LICENSEcommand ins ta l l a new l i cense.

LOG ADMIN

Authorizes the ALTER SYSTEMLOGGING [ON|OFF] commands toenable or disable the log flushmechanism.

MONITOR ADMINAuthorizes the ALTER SYSTEMcommands for EVENTs .

OPTIMIZER ADMIN

Authorizes the ALTER SYSTEMcommands concerning SQL PLANCACHE and ALTER SYSTEM UPDATESTATISTICS commands , whichinfluence the behavior of the queryoptimizer.

RESOURCE ADMIN

This privi lege authorizes commandsconcerning sys tem resources , forexample ALTER SYSTEM RECLAIMDATAVOLUME and ALTER SYSTEMRESET MONITORING VIEW. It a l soauthorizes many of the commandsava i lable in the ManagementConsole.

ROLE ADMIN

This privi lege authorizes the creationand deletion of roles us ing theCREATE ROLE and DROP ROLEcommands . It a l so authorizes thegranting and revocation of rolesus ing the GRANT and REVOKEcommands .

Activated roles , meaning roles whosecreator i s the pre-defined user _SYS_REPO,can nei ther be granted to other roles orusers nor dropped di rectly. Not even us ershaving ROLE ADMIN privi lege are able todo so. Please check documentationconcerning activated objects .

SAVEPOINT ADMINAuthorizes the execution of asavepoint process us ing the ALTER


160 2012-12-19

SYSTEM SAVEPOINT command.

SCENARIO ADMINAuthorizes a l l ca lculation scenario-related activi ties (includingcreation).

SERVICE ADMINAuthorizes the ALTER SYSTEM[START|CANCEL|RECONFIGURE]commands .

This privi lege i s for adminis tering sys temservices of the databas e

SESSION ADMIN

authorizes the ALTER SYSTEMcommands concerning sess ions tos top or disconnect a user sess ion orto change sess ion variables .

STRUCTUREDPRIVILEGEADMIN

Authorizes the creation, reactivation,and dropping of s tructuredprivi leges .

TRACE ADMIN

Authorizes the ALTER SYSTEM[CLEAR|REMOVE] TRACES commandsfor operations on database tracefi les and authorizes changing tracesys tem settings .

TRUST ADMINAuthorizes commands to update thetrust s tore.

USER ADMIN

Authorizes the creation andmodi fication of users us ing theCREATE USER, ALTER USER, and DROPUSER commands .

VERSION ADMIN

Authorizes the ALTER SYSTEM RECLAIMVERSION SPACE command of themulti -vers ion concurrency control(MVCC) mechanism.

<identi fier>.<identi fier>

Components of the SAP HANAdatabase can create new sys temprivi leges . These privi leges use thecomponent-name as fi rs t identi fierof the sys tem privi lege and thecomponent-privi lege-name as thesecond identi fier.

<schema_privilege> ::= CREATE ANY | DEBUG | DELETE | DROP | EXECUTE | INDEX | INSERT | SELECT | TRIGGER | UPDATE

Schema privileges are used to restrict the access and modifications on a schema and the objects stored inthis schema. The following schema privileges are defined:

CREATE ANY This privilege allows the creation of all kinds of objects, in particular, tables, views, sequences, synonyms,SQLScript functions, or database procedures in a schema.

DELETE, DROP, EXECUTE, INDEX, INSERT, SELECT, UPDATE The specified privilege is granted on every object stored in the specified schema currently and in future. Fordetailed description of the privileges please see the part describing object privileges below and please checkbelow, which privileges are applicable for which kind of object.

<object_privilege> ::= ALL PRIVILEGES | ALTER | DEBUG | DELETE | DROP | EXECUTE | INDEX | INSERT | SELECT | REFERENCES | TRIGGER | UPDATE | <identifier>.<identifier>

Object privileges are used to restrict the access and modifications on database objects. Database objects aretables, views, sequences, procedures, etc. The table below describes the supported object privileges.


161 2012-12-19

Object PrivilegeCommandTypes

Privilege Description Comment

ALL PRIVILEGESDDL &DML

This privi lege i s a col lection of a l lData Defini tion Language(DDL)and Data ManipulationLanguage(DML) privi leges that thegrantor currently possesses and i sa l lowed to grant further. Theprivi lege i t grants i s speci fi c to theparticular object being actedupon.

This privi lege col lection i sdynamica l ly eva luated for thegiven grantor and object. ALLPRIVILEGES i s appl icable to atable or view

ALTER DDLAuthorizes the ALTER command forthe object

DEBUG DML

Authorizes debug-functiona l i ty forthe procedure or ca lculation viewor for the procedures andca lculation views of a schema.

DELETE DMLAuthorizes the DELETE andTRUNCATE commands for theobject

DROP DDLAuthorizes the DROP commandsfor the object

EXECUTE DML

Authorizes the execution of anSQLScript function or a databaseprocedure us ing the CALLS or CALLcommand respectively.

INDEX DDLAuthorizes the creation,modi fication or dropping ofindexes for the object

INSERT DMLAuthorizes the INSERT commandfor the object.

The INSERT and UPDATEprivi lege are both required onthe object to a l low the REPLACEand UPSERT commands to beused.

REFERENCES DDLAuthorizes the usage of a l l tablesin this schema or this table in aforeign key defini tion.

SELECT DMLAuthorizes the SELECT commandfor this object or the usage of asequence.

TRIGGER DDL

Authorizes the CREATE TRIGGER /DROP TRIGGER command for thespeci fied table or the tables inthe speci fied schema.

UPDATE DMLAuthorizes the UPDATE commandfor that object.

he INSERT and UPDATE privi legeare both required on the objectto a l low the REPLACE andUPSERT commands to be used.The UPDATE privi lege i s a l sorequired to perform del tamerges of column s tore tablesus ing the MERGE DELTAcommand.

<identi fier>.<identi fier> DDL

Components of the SAP HANAdatabase can create new sys temprivi leges . These privi leges usethe component-name as fi rs tidenti fier of the sys tem privi legeand the component-privi lege-name as the second identi fier.

Not all object privileges are applicable to all kinds of database objects. For details of which object types allowwhich privilege to be used please see the table below.


162 2012-12-19

Privilege Schema Table View Sequence Function/Procedure

ALL PRIVILEGES --- YES YES --- ---

ALTER YES YES --- --- YES

CREATE ANY YES --- --- --- ---

DEBUG YES --- YES --- YES

DELETE YES YES YES --- ---

DROP YES YES YES YES YES

EXECUTE YES --- --- --- YES

INDEX YES YES --- --- ---

INSERT YES YES YES --- ---

REFERENCES YES YES --- --- ---

SELECT YES YES YES YES ---

TRIGGER YES YES --- --- ---

UPDATE YES YES YES --- ---

DELETE, INSERT and UPDATE on views are valid for updatable views only. Updatable views have the followingcharacteristics:

Does not use a joinDoes not contain a UNIONDoes not use aggregation

DEBUG is only valid for calculation views. For synonyms, the same restrictions apply to the synonym as they would do for the object the synonymrepresents.

<object_name> ::= <table_name> | <view_name> | <sequence_name> | <procedure_name> | <synonym_name>

Object privileges are used to restrict the access and modifications on database objects like tables, views,sequences, procedures and synonyms.

<grantee> :: = <user_name> | <role_name>

The grantee can be a user or a role. In case a privilege or role is granted to a role, then all user being grantedthat role, will have the specified privilege or role.

A role is a named collection of privileges and can be granted to either a user or a role. If you want to allow several database users to perform the same actions, you should create a role, grant theneeded privileges to this role, and finally grant the role to the different database users. When granting roles to roles, a tree of roles can be build. When granting a role (R) to a role or user (G), G willhave all privileges directly granted to R and all privileges granted to roles which had been granted to R.


The grantee username.


The grantee role name.

<schema_name> ::= <identifier>

The schema containing the database object to be acted upon.

<table_name> ::= [<schema_name>.]<identifier>

The database table object name.

<view_name> ::= [<schema_name>.]<identifier>


163 2012-12-19

The database view object name.

<sequence_name> ::= [<schema_name>.]<identifier>

The sequence object name.

<procedure_name> ::= [<schema_name>.]<identifier>

The procedure object name.

<synonym_name> ::= [<schema_name>.]<identifier>

The synonym name.

<privilege_name> ::= <identifier>

The name of the privilege to be granted.

WITH ADMIN OPTION and WITH GRANT OPTIONSpecifies that the granted privileges can be granted further by the specified user or by those user having thespecified role.

With GRANT STRUCTURED PRIVILEGE <privilege_name> A previously defined analytical privilege based on a generic structured privilege is granted to a user or role.This analytical privileges is used to restrict the access for read operations to certain data in Analytic,Attribute, and Calculation Views by filtering the attribute values.

Description

GRANT is used to grant privileges and structured privileges to users and roles. GRANT is also used to grantroles to users and other roles.

The specified users, roles, objects, and structured privileges have to exist before they can be used in theGRANT command. In order to use the GRANT command to grant privileges to other users and roles, a user must have theprivilege and also the permissions required to grant that privilege.

A user can not grant a privilege to themselves. User SYSTEM has at all system privileges and the role PUBLIC. All other users have the role PUBLIC. Theseprivileges and roles cannot be revoked.

Although the SYSTEM user has many privileges, it cannot select or change data in other user's tables unlessthis privilege has been explicitly granted.

All users have the privilege to create objects in their own default schema. The default schema for a useralways has the same schema name as the user.

For tables created by users, users have all privileges and may grant all privilges further to other users androles.

For objects which are dependent on other objects, like views being dependent on tables, it can occur theowner of the dependent object does not have a complete set of privileges. This can occur if the user do nothave the privileges on the underlying objects on which their object depends.

Users can have privileges on an object, but may not have sufficient priviliges to grant them to other users androles.


USERS: shows all users, their creator, creation date and some info about their current states. ROLES: shows all roles, their creator and creation date. GRANTED_ROLES: shows which roles are granted to which user or role. GRANTED_PRIVILEGES: shows which privileges are granted to which user or role.

Examples

You create a schema called my_schema.


164 2012-12-19

CREATE SCHEMA my_schema;

You create a table named work_done in the my_schema schema.

CREATE TABLE my_schema.work_done (t TIMESTAMP, user NVARCHAR (256), work_done VARCHAR (256));

You create a new user named worker with password His_Password_1.

CREATE USER worker PASSWORD His_Password_1;

You create role called role_for_work_on_my_schema.


You grant the SELECT on any object privilege in my_schema to the role role_for_work_on_my_schema.

GRANT SELECT ON SCHEMA my_schema TO role_for_work_on_my_schema;

You grant the INSERT privilege for the work_done table to the role role_for_work_on_my_schema.

GRANT INSERT ON my_schema.work_done TO role_for_work_on_my_schema;

You grant the role_for_work_on_my_schema role to the worker user.

GRANT role_for_work_on_my_schema TO worker WITH ADMIN OPTION;

You grant DELETE privilige for this table to the worker user.

GRANT DELETE ON my_schema.work_done TO worker;

You grant the worker user the privilege to create any kind of object in the my_schema schema.

GRANT CREATE ANY ON SCHEMA my_schema TO worker;

The result of the above commands is that the worker user has the privilege to SELECT all tables and views inschema my_schema, to INSERT into and DELETE from table my_schema.work_done and to create objects inschema my_schema. Additionally the worker user is allowed to grant DELETE on the tablemyschema.work_done to other users and roles.

You grant the privileges INIFILE ADMIN and TRACE ADMIN to the user worker. You grant these privilegesalong with the permission for the worker user to grant them further.

GRANT INIFILE ADMIN, TRACE ADMIN TO worker WITH ADMIN OPTION;

REVOKE

Syntax

REVOKE <system_privilege>,... FROM <grantee> | REVOKE <schema_privilege>,... ON SCHEMA <schema_name> FROM <grantee> | REVOKE <object_privilege>,... ON <object_name> FROM <grantee> | REVOKE <role_name>,... FROM <grantee> | REVOKE STRUCTURED PRIVILEGE <privilege_name> FROM <grantee> | REVOKE APPLICATION PRIVILEGE <privilege_name> FROM <grantee>

Syntax Elements

For the definition of syntax elements, please refer to GRANT.

Description

The REVOKE statement revokes the specified roles or structured privileges or the privileges for the specifiedobjects from the specified users or roles. Only users having granted a privilege can revoke that privilege. This is true even for users having ROLEADMIN and the revoke of roles.User SYSTEM has at least all system privileges and the role PUBLIC. All other users have the role PUBLIC.These privileges and roles can not be revoked from them.


165 2012-12-19

If a user was granted a role, it is not possible to revoke some of the privileges belonging to that role. In thiscase, the whole role has to be revoked and those privileges needed by that user granted to him. Revoking a privilege or role may lead to making some views inaccessible or making procedures notexecutable any more. This will occur if a view or procedures depends on the privilege revoked or on one ofthe privileges the role had. Revoking a privilege which had been granted with WITH GRANT OPTION or with WITH ADMIN OPTION willresult in revoking that privilege not only from the user specified in the command, but from all the users androles having been granted that privilege by this user directly and indirectly.

As privileges can be granted to one user or role by different users, revoking of this privilege by one user doesnot necessarily mean, that this user will loose that privilege. For the description of syntax elements andprivileges, please refer to GRANT.


USERS: shows all users, their creator, creation date and some info about their current states. ROLES: shows all roles, their creator and creation date. GRANTED_ROLES: shows which roles are granted to which user or role. GRANTED_PRIVILEGES: shows which privileges are granted to which user or role.

Example

You create new user named worker.

CREATE USER worker PASSWORD His_Password_1;

You create new role named role_for_work_on_my_schema.


You create new schema named my_schema.

CREATE SCHEMA my_schema OWNED BY system;

You create new table named work_done in the schema.

CREATE TABLE my_schema.work_done (t TIMESTAMP, user NVARCHAR (256), work_done VARCHAR (256) );

You grant the SELECT on any object privilege in my_schema to the role role_for_work_on_my_schema.

GRANT SELECT ON SCHEMA my_schema TO role_for_work_on_my_schema;

You grant the INSERT privilege for the work_done table to the role role_for_work_on_my_schema.

GRANT INSERT ON my_schema.work_done TO role_for_work_on_my_schema;

You grant the role_for_work_on_my_schema role to the worker user.

GRANT role_for_work_on_my_schema TO worker;

You grant TRACE ADMIN privilege for the worker user.

GRANT TRACE ADMIN TO worker WITH ADMIN OPTION;

You grant DELETE from work_done table to the worker user.

GRANT DELETE ON my_schema.work_done TO worker WITH GRANT OPTION;

You revoke from the role role_for_work_on_my_schema the privilege to select from my_schema.

REVOKE SELECT ON SCHEMA my_schema FROM role_for_work_on_my_schema;

You revoke TRACE ADMIN privilege from the worker.

REVOKE TRACE ADMIN FROM worker;

Data Import Export Statements


166 2012-12-19

EXPORT

Syntax

EXPORT <object_name_list> AS <export_format> INTO <path> [WITH <export_option_list>]

Syntax Elements

WITH <export_option_list>:

EXPORT options can be passed in using "WITH" clause.

<object_name_list> ::= <OBJECT_NAME>,... | ALL <export_import_format> ::= BINARY | CSV <path> ::= 'FULL_PATH' <export_option_list> ::= <export_option> | <export_option_list> <export_option> <export_option> ::= REPLACE | CATALOG ONLY | NO DEPENDENCIES | SCRAMBLE [BY <password>] | THREADS <number_of_threads>

Description

The EXPORT command exports tables, views, column views, synonyms, sequences, or procedures in thespecified format as BINARY or CSV. Data for temporary tables and "no logging" tables cannot be exportedusing EXPORT.

OBJECT_NAMESQL name of the object (table, view, etc) to export. To export all objects from all schemas you should use thekeyword ALL. If you want to export objects of a specific schema, you should use the schema name with theasterisk, e.g. "SYSTEM"."*"

BINARYTable data will be exported in internal binary format. Exporting in this format is orders of magnitude fasterthen exporting the same table in CSV format. Only column tables can be exported in binary format. The rowtables are always exported in CSV format, even if BINARY format is specified.

CSVTable data will be exported in CSV format. The exported data can be imported only in any other database.Additionaly, the exported data can be scramled. Both column and row tables can be exported in CSV format.

FULL_PATHThe server path to export to.

Note:When using a distributed system, the FULL_PATH must point to a shared disk! For security reasons,the path may not contain symbolic links and may not point inside the database instance folder,except its 'backup' and 'work' subfolders. Examples for valid export path (assuming the databaseinstance is located at /usr/sap/HDB/HDB00):

'/tmp'<br> '/usr/sap/HDB/HDB00/backup'<br> '/usr/sap/HDB/HDB00/work'<br>

REPLACEWith REPLACE option, previously exported data will be removed and newly exported data will be saved. IfREPLACE option is not specified, an error will be thrown if there exists a previously exported data already inthe specified directory.

CATALOG ONLYWith CATALOG ONLY option, only database catalog will be exported without data.

NO DEPENDENCIESWith NO DEPENDENCIES option, it will not export underlying objects of an object to be exported.

SCRAMBLEExport in CSV format has an additional option to scramble sensitive customer data using SCRAMBLE [BY'<password>']. When the optional password is not specified, a default scramble password is used. Onlycharacter string data can be scrambled. When imported, scrambled data will be imported as scrambled so


167 2012-12-19

that end users cannot read the data and it is not possible to unscramble the data.

THREADSIndicates the number of threads that can be used for concurrent export.

Number of threads to use

The given number of THREADS specifies how many objects will be exported in parallel (default is 1).Increasing number of threads may reduce export time, but also can negatively affect the systemperformance.

Following should be considered:

For a single table THREADS has no effect.For a view or procedure, 2 or more threads should be used (up to number of objects it depends on)For a whole schema consider using more that 10 threads (up to number of cores in the system)For a whole BW / ERP system (ALL keyword) with tens of thousands of tables, a large number ofthreads can be reasonable (up to 256)


You can monitor the progress of the export using M_EXPORT_BINARY_STATUS system views.

You can abort the export session using the session ID from the corresponding view in the following command:

ALTER SYSTEM CANCEL [WORK IN] SESSION 'sessionId'

The detailed result of the export is stored in session-local temporary tables. #EXPORT_RESULT.


You create new schema named MY_SCHEMA.

CREATE SCHEMA MY_SCHEMA;

You export the information from MY_SCHEMA in CSV format with the REPLACE and SCRAMBLE optionsenabled.

EXPORT "MY_SCHEMA"."*" AS CSV INTO '/tmp' WITH REPLACE SCRAMBLE THREADS 10

IMPORT

Syntax

IMPORT <object_name_list> [AS <import_format>] FROM <path> [WITH <import_option_list>]

Syntax Elements

WITH <import_option_list>:

IMPORT options can be passed in using "WITH" clause.

<object_name_list> ::= <object_name>,... | ALL <import_format> ::= BINARY | CSV <path> ::= 'FULL_PATH' <import_option_list> ::= <import_option> | <import_option_list> <import_option> <import_option> ::= REPLACE | CATALOG ONLY | NO DEPENDENCIES | THREADS <number_of_threads>

Description

The IMPORT statement imports tables, views, column views, synonyms, sequences, or procedures. Data fortemporary tables and "no logging" tables cannot be imported using IMPORT command.

OBJECT_NAMESQL name of the object (table, view, etc) to import. To import all objects from the import path you should use


168 2012-12-19

the keyword ALL. If you want to import objects of a specific schema, you should use the schema name withthe asterisk, e.g. "SYSTEM"."*"

BINARY | CSVFormat specification may be ommited during the import, since the format will be auto-detected duringimport. The data will be imported in the same format in which it was exported.

FULL_PATHThe server path to import from.

Note:When using a distributed system, the FULL_PATH must point to a shared disk! Examples for validimport path (assuming the database instance is located at /usr/sap/HDB/HDB00):

'/tmp'<br> '/usr/sap/HDB/HDB00/backup'<br> '/usr/sap/HDB/HDB00/work'<br>

REPLACEWith REPLACE option, the specified table will be dropped and created and data will be imported. If REPLACEoption is not specified, an error will be thrown if there exists a table with the same name in the sameschema.

CATALOG ONLYWith CATALOG ONLY option, only database catalog will be imported without data.

NO DEPENDENCIESWith NO DEPENDENCIES option, it will not import underlying objects of an object to be imported.

THREADSIndicates the number of threads that can be used for concurrent import.


The given number of THREADS specifies how many objects will be imported in parallel (default is 1).Increasing number of threads may reduce import time, but also can negatively affect the systemperformance.

Following should be considered:

For a single table THREADS has no effect.For a view or procedure, 2 or more threads should be used (up to number of objects it depends on)For a whole schema consider using more that 10 threads (up to number of cores in the system)For a whole BW / ERP system (ALL keyword) with tens of thousands of tables, a large number ofthreads can be reasonable (up to 256)


You can monitor the progress of the import using M_IMPORT_BINARY_STATUS system views.

Also, you can abort the import session using the connection ID from the corresponding view in the followingcommand:

ALTER SYSTEM CANCEL [WORK IN] SESSION 'sessionId'

The detailed result of the import is stored in session-local temporary tables. #IMPORT_RESULT.

Example

IMPORT "SCHEMA"."*" AS CSV FROM '/tmp' WITH REPLACE THREADS 10

IMPORT FROM

Syntax

IMPORT FROM [<file_type>] <file_path> [INTO <table_name>] [WITH <import_from_option_list>]

Syntax Elements

WITH <import_from_option_list>:


169 2012-12-19

IMPORT FROM options can be passed in using "WITH" clause.

<file_type> ::= CSV FILE | CONTROL FILE

The type of the file to be imported. You can specify either Comma-separated values or control file formats.For details of CSV and control file formats please see the Examples.

<file_path> ::= <string_literal>

The complete path of the file and filename of the file to import.


The target table name, with optional schema name, where the imported data will be stored.

WITH <import_from_option_list> <import_from_option_list> ::= <import_from_option> ...

The optional WITH option list can have one or many of the options stated below.

<import_from_option> ::= THREADS <number_of_threads> | BATCH <number_of_records_of_each_commit> | TABLE LOCK | NO TYPE CHECK | SKIP FIRST <number_of_rows_to_skip> ROW | COLUMN LIST IN FIRST ROW | COLUMN LIST ( <column_name_list> ) | RECORD DELIMITED BY <string_for_record_delimiter> | FIELD DELIMITED BY <string_for_field_delimiter> | OPTIONALLY ENCLOSED BY <character_for_optional_enclosure> | DATE FORMAT <string_for_date_format> | TIME FORMAT <string_for_time_format> | TIMESTAMP FORMAT <string_for_timestamp_format> | ERROR LOG <file_path_of_error_log>

Import option parameters are detailed below.

THREADS <number_of_threads> <number_of_threads> ::= <integer_literal>

The number of threads that can be used for concurrent import. The default value is 1 and maximum allowedis 256.

BATCH <number_of_records_of_each_commit> <number_of_records_of_each_commit> ::= <integer_literal>

The number of records to be inserted in each commit.

Note:THREADS and BATCH can be used to achieve high loading performance by enabling parallel loadingand also by committing many records at once. In general, for column tables, a good setting to use is10 parallel loading threads, with a commit frequency of 10.000 records or greater.

TABLE LOCK

Locks table for fast import into column store tables. If NO TYPE CHECK option is specified, then the record isinserted without checking the type of each field.

SKIP FIRST <number_of_rows_to_skip> ROW <number_of_rows_to_skip> ::= <integer_literal>

Skips to insert first n rows.

COLUMN LIST IN FIRST ROW

Indicates the column list is stored in the first row of the CSV import file.

COLUMN LIST ( <column_name_list> ) <column_name_list> ::= <column_name> [{, <column_name>}...] <column_name> ::= <string_literal>


170 2012-12-19

The column list for the data being imported. The name list has one or more column names. The ordering ofthe column names should match the order of the column data in the CSV file and the columns in the targettable.

RECORD DELIMITED BY <string_for_record_delimiter> <string_for_record_delimiter> ::= <string_literal>

The record delimiter used in the CSV file being imported.

FIELD DELIMITED BY '<string_for_field_delimiter>' <string_for_field_delimiter> ::= <string_literal>

Indicates the field delimiter of the CSV file

OPTIONALLY ENCLOSED BY <character_for_optional_enclosure> <character_for_optional_enclosure> ::= <character_literal>

Indicates the optional enclosure character used to delimit field data

DATE FORMAT <string_for_date_format> <string_for_date_format> ::= <string_literal>

The format that date strings are encoded with in the import data.Y : year , MM : month , MON : name of month, DD : day For example:'YYYYMMDD' : 20120520 , 'YYYY-MM-DD' : 2012-05-20 , 'YYYY-MON-DD' : 2012-MAY-20

TIME FORMAT <string_for_time_format> <string_for_time_format> ::= <string_literal>

The format that time strings are encoded with in the import data.HH24 : hour , MI : minute , SS : secondFor example:'HH24MISS' : 143025 , 'HH24:MI:SS' : 14:30:25

TIMESTAMP FORMAT <string_for_timestamp_format> <string_for_timestamp_format> ::= <string_literal>

The format that timestamp strings are encoded with in the import data.For example:'YYYY-MM-DD HH24:MI:SS' : 2012-05-20 14:30:25

ERROR LOG <file_path_of_error_log> <file_path_of_error_log> ::= <string_literal>

You can store a log of errors that occur during import. Please ensure the file path you use is writeable by thedatabase.

Note:All <string_literal>s in the <import_from_option> support UTF-8 except surrogate-pair encoding.

Description

The IMPORT FROM statement imports external data from a file into an existing table.

Examples

Example 1 -Importing CSV data

You create a table MYTABLE to store the imported data.

CREATE TABLE MYTABLE ( A INT, B VARCHAR(10), C DATE, D TIME, E DECIMAL );

You create a CSV text file /data/data.csv and add the following contents.

1,"DATA1","2012-05-20","14:30:25",123456 2,"DATA2","2012-05-21","15:30:25",234567 3,"DATA3","2012-05-22","16:30:25",345678 4,"DATA4","2012-05-23","17:30:25",456789


171 2012-12-19

You execute the following command to import the data.

IMPORT FROM CSV FILE '/data/data.csv' INTO "MYTABLE" WITH RECORD DELIMITED BY '\n' FIELD DELIMITED BY ',';

Example 2 - Importing using a control file

In the example below you import the CSV data from Example 1 using a control file:

You can create a control file /data/data.ctl and add the contents shown below to the file.

IMPORT DATA INTO TABLE "MYTABLE" FROM '/data/data.csv' RECORD DELIMITED BY '\n' FIELD DELIMITED BY ',' OPTIONALLY ENCLOSUED BY '"' ERROR LOG '/data/data.err'

You execute the following command to import the data using the control file.

IMPORT FROM CONTROL FILE '/data/data.ctl';

Example 3 - Import using date formats

In the example below the date format is of the CSV import data is different to the default date format 'YYYY-MM-DD'. In this import data the date format used is 'MM-DD-YYYY'. You create a CSV text file /data/data_different_date.csv and add the following contents.

1,"DATA1","05-20-2012","14:30:25",123456 2,"DATA2","05-21-2012","15:30:25",234567 3,"DATA3","05-22-2012","16:30:25",345678 4,"DATA4","05-23-2012","17:30:25",456789

You execute the following command to import the data.

IMPORT FROM CSV FILE '/data/data_different_date.csv' INTO "MYTABLE" WITH RECORD DELIMITED BY '\n' FIELD DELIMITED BY ',' DATE FORMAT 'MM-DD-YYYY';

Example 4 - Import using COLUMN LIST

You create a table called COLLIST to store the imported data.

CREATE TABLE COLLIST ( A INT, B VARCHAR(10), c DATE, D DECIMAL );

You create a CSV text file '/data/data_col_list.csv' and add the following contents.

1,"DATA1","2012-05-20",123456 2,"DATA2","2012-05-21",234567 3,"DATA3","2012-05-22",345678 4,"DATA4","2012-05-23",456789

You execute the following commands to import the data using a column list.

IMPORT FROM CSV FILE '/data/data_col_list.csv' INTO "COLLIST" WITH RECORD DELIMITED BY '\n' FIELD DELIMITED BY ',' COLUMN LIST ("A", "B", "D");

IMPORT SCAN

Syntax

IMPORT SCAN <path>

Syntax Elements

<path> ::= <string_literal>

The path of the directory to be scanned for import data.

Description


172 2012-12-19

The IMPORT SCAN statement scans the path and fills the session-local temporary table#IMPORT_SCAN_RESULT with the contents in the path.

Example

You scan the path '/data/import_path' for import data.

IMPORT SCAN '/data/import_path'

You select the data in the #IMPORT_SCAN_RESULT table to view the result of the scan.

SELECT * FROM #IMPORT_SCAN_RESULT;

Procedural Statements

ALTER PROCEDURE RECOMPILE

Syntax

ALTER PROCEDURE <proc_name> RECOMPILE [WITH PLAN]

Syntax Elements

<proc_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The identifier of the procedure to be altered, with optional schema name.

WITH PLAN

Specifies that internal debug information should be created during execution of the procedure.

Description

The ALTER PROCEDURE RECOMPILE statement manually triggers a recompilation of a procedure bygenerating an updated execution plan. For production code a procedure should be compiled without theWITH PLAN option to avoid overhead during compilation and execution of the procedure.

Example

You trigger the recompilation of the my_proc procedure to produce debugging information.

ALTER PROCEDURE my_proc RECOMPILE WITH PLAN;

CALL

Syntax

CALL <proc_name> (<param_list>) [WITH OVERVIEW] [IN DEBUG MODE]

Syntax Elements


The identifier of the procedure to be called, with optional schema name.

<param_list> ::= <proc_param>[{, <proc_param>}...]

Specifies one or more procedure parameters.

<proc_param> ::= <identifier> | <string_literal> | <unsigned_integer> | <signed_integer>| <signed_numeric_literal> | <unsigned_numeric_literal>

Procedure parameters. For more information on these data types please see BNF Lowest TermsRepresentations and Data Types.


173 2012-12-19

Parameters passed to a procedure are scalar constants and can be passed either as IN, OUT or INOUTparameters. Scalar parameters are assumed to be NOT NULL. Arguments for IN parameters of table typecan either be physical tables or views. The actual value passed for tabular OUT parameters must be`?`.

WITH OVERVIEW

Defines that the result of a procedure call will be stored directly into a physical table.

Calling a procedure WITH OVERVIW will return one result set that holds the information of which tablecontains the result of a particular table’s output variable. Scalar outputs will be represented as temporarytables with only one cell. When you pass existing tables to the output parameters WITH OVERVIEW will insertthe result set tuples of the procedure into the provided tables. When you pass NULL to the outputparameters, temporary tables holding the result sets will be generated. These tables will be droppedautomatically once the database session is closed.

IN DEBUG MODE

When specified additional debug information will be created during the execution of the procedure. Thisinformation can be used to debug the instantiation of the internal execution plan of the procedure.

Note:that this type of call implies additional runtime overhead and so it should only be used for debuggingpurposes.

Description

Calls a procedure defined with CREATE PROCEDURE.

CALL conceptually returns list of result sets with one entry for every tabular result. An iterator can be used toiterate over these results sets. For each result set you can iterate over the result table in the same way as forquery results. SQL statements that are not assigned to any table variable in the procedure body will be addedas result sets at the end of the list of result sets. The type of the result structures will be determined duringcompilation time but will not be visible in the signature of the procedure.

CALL when executed by the client the syntax behaves in a way consistent with the SQL standard semantics,e.g. Java clients can call a procedure using a JDBC CallableStatement. Scalar output variables will be a scalarvalue that can be retrieved from the callable statement directly.

Note:Unquoted identifiers are implicitly treated as upper case. Quoting identifiers will respectcapitalization and allow for using white spaces which are normally not allowed in SQL identifiers.

Examples

You call the getOutput procedure in debug mode.

CALL getOutput (1000, 'EUR', NULL, NULL) IN DEBUG MODE;

You call the procedure getOutput using the WITH OVERVIEW option.

CALL getOutput(1000, 'EUR', ot_publishers, NULL) WITH OVERVIEW;

CREATE FUNCTION

Syntax

CREATE FUNCTION <func_name> [(<parameter_clause>)] RETURNS <return_type> [LANGUAGE <lang>] [SQL SECURITY <mode>] AS <local_scalar_variables> BEGIN <function_code> END

Syntax Elements

<func_name > ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>


174 2012-12-19

The identifier of the function to be created, with optional schema name.

<parameter_clause> ::= <parameter> [{,<parameter>}…]

The input parameters of the function.

<parameter> ::= [IN] <param_name> <datatype>

A function parameter with associated data type.

<param_name> ::= <identifier>

The variable name for a parameter.

<datatype> ::= <sql_type> | <table_type>

The input parameters of a function can have any of the primitive SQL types or a table type.

<sql_type> ::= BIGINT | INTEGER | SMALLINT | TINYINT | DATE | TIME | TIMESTAMP | SMALLDECIMAL | DECIMAL | REAL | DOUBLE

The data type of the variable. Please see Data Types.

<table_type> ::= <identifier>

A table type previously defined with the CREATE TYPE command. Please see CREATE TYPE.

<return_type> := <sql_type> | <return_table_type>

The returned results of a function can have any of the primitive SQL types or a return table type.

<return_table_type> := TABLE (<ret_column_list>)

Defines the structure of the returned table data.

<ret_column_list> := <ret_column_elem>[{, <ret_column_elem>}]

Defines the list of columns returned from the function.

<ret_column_elem> := <column_name> <sql_type> <column_name> := <identifier>

Defines the name of column element with associated datatype.

LANGUAGE <lang> <lang> ::= SQLSCRIPT

Default: SQLSCRIPTDefines the programming language used in the function.

Note:Only SQLScript functions can currently be defined.

SQL SECURITY <mode> <mode> ::= DEFINER | INVOKER

Default: DEFINERSpecifies the security mode of the function.

DEFINER

Specifies that the execution of the function is performed with the privileges of the definer of the function.

INVOKER

Specifies that the execution of the function is performed with the privileges of the invoker of the function.

<local_scalar_variables> ::= {<local_var>;}...


175 2012-12-19

Defines local scalar variables for the function.

<local_var> ::= <variable_name> <datatype> <variable_name> ::= <identifier>

Defines a local function variable with associated scalar type. Please see Data Types.

<function_code> := !! function statements.

Defines the main body of the function.

Description

The CREATE FUNCTION statement creates a function read only function which is free of side-effects. That isto say that neither DDL or DML statements (INSERT, UPDATE, and DELETE) are allowed in the function body.Also other functions or procedures selected/called from the body of the function must be read-only.

Example

You create an SQLScript function with the following definition.

CREATE FUNCTION scale (val INT) RETURNS TABLE (a INT, b INT) LANGUAGE SQLSCRIPT AS BEGIN RETURN SELECT a, :val * b AS b FROM mytab; END;

You use the scale function with the following example select statements.

select * from scale(10); select * from scale(10) as a, scale(10) as b where a.a = b.a

CREATE PROCEDURE

Syntax

CREATE PROCEDURE <proc_name> [(<parameter_clause>)] [LANGUAGE <lang>] [SQL SECURITY <mode>] [READS SQL DATA [WITH RESULT VIEW <view_name>]] AS <local_scalar_variables> BEGIN <procedure_code> END

Syntax Elements

<proc_name > ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The identifier of the procedure to be created, with optional schema name.

<parameter_clause> ::= <parameter> [{,<parameter>}…]

The input and output parameters of the procedure.

<parameter> ::= [<param_inout>] <param_name> <datatype>

A procedure parameter with associated data type.

<param_inout> ::= IN|OUT|INOUT

Default: INEach parameter is marked using the keywords IN/OUT/INOUT. Input and output parameters must beexplicitly typed (i.e. no un-typed tables are supported).

<param_name> ::= <identifier>

The variable name for a parameter.



176 2012-12-19


The input and output parameters of a procedure can have any of the primitive SQL types or a table type.INOUT parameters can only be of scalar type.

<sql_type> ::= DATE | TIME | SECONDDATE | TIMESTAMP | TINYINT | SMALLINT | INTEGER | BIGINT | SMALLDECIMAL | DECIMAL | REAL | DOUBLE | VARCHAR | NVARCHAR | ALPHANUM | SHORTTEXT | VARBINARY | BLOB | CLOB | NCLOB | TEXT

The data type of the variable. Please see Data Types.

<table_type> ::= <identifier>

A table type previously defined with the CREATE TYPE command. Please see CREATE TYPE.

LANGUAGE <lang> <lang> ::= SQLSCRIPT | R

Default: SQLSCRIPTDefines the programming language used in the procedure. It is good practice to define the language in allprocedure definitions.

SQL SECURITY <mode> <mode> ::= DEFINER | INVOKER

Default: DEFINERSpecifies the security mode of the procedure.

DEFINER

Specifies that the execution of the procedure is performed with the privileges of the definer of the procedure.

INVOKER

Specifies that the execution of the procedure is performed with the privileges of the invoker of theprocedure.

READS SQL DATA

Marks the procedure as being read-only, side-effect free i.e. the procedure does not make modifications tothe database data or its structure. This means that the procedure does not contain DDL or DML statements,and that the procedure only calls other read-only proceedures. The advantage of using this parameter is thatcertain optimizations are available for read-only procedures.

WITH RESULT VIEW <view_name> <view_name> ::= <identifier>

Specifies the result view to be used as the output of a read-only procedure.

When a result view is defined for a procedure, it can be called by an SQL statement in the same way as atable or view. Please see Example 2 - Using a result view below.

<local_scalar_variables> ::= {<local_var>;}...

Defines local scalar variables for the procedure.

<local_var> ::= <variable_name> <datatype> <variable_name> ::= <identifier>

Defines a local procedure variable with associated scalar type. Please see Data Types.

<procedure_code> := !! Procedure statements.

Defines the main body of the procedure according to the programming language selected.For more information about SQLScript please see the "SQLScript reference". For more information on the R-Language please see the "R-Language reference". Both reference documents are available from the SAPHANA Appliance page.

Description


177 2012-12-19


The CREATE PROCEDURE statement creates a procedure using the specified programming language <lang>.

Examples

Example 1 - Creating an SQL Procedure

You create an SQLScript procedure with the following definition.

CREATE PROCEDURE orchestrationProc LANGUAGE SQLSCRIPT AS v_id BIGINT; v_name VARCHAR(30); v_pmnt BIGINT; v_msg VARCHAR(200); CURSOR c_cursor1 (p_payment BIGINT) FOR SELECT id, name, payment FROM control_tab WHERE payment > :p_payment ORDER BY id ASC; BEGIN CALL init_proc(); OPEN c_cursor1(250000); FETCH c_cursor1 INTO v_id, v_name, v_pmnt; v_msg := :v_name || ' (id ' || :v_id || ') earns ' || :v_pmnt || ' $.'; CALL ins_msg_proc(:v_msg); CLOSE c_cursor1; END;

The procedure features a number of imperative constructs including the use of a cursor (with associatedstate) and local scalar variables with assignments.

Example 2 - Using a result view

You create a procedure using a result view ProcView to return its results.

CREATE PROCEDURE ProcWithResultView(IN id INT, OUT o1 CUSTOMER) LANGUAGE SQLSCRIPT READS SQL DATA WITH RESULT VIEW ProcView AS BEGIN o1 = SELECT * FROM CUSTOMER WHERE CUST_ID = :id; END;

You call this procedure from an SQL statement as follows.

SELECT * FROM ProcView WITH PARAMETERS ('placeholder' = ('$$id$$', '5'));

CREATE TYPE

Syntax

CREATE TYPE <type_name> AS TABLE (<column_definition>[{,<column_definition>}...])

Syntax Elements

<type_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

Identifies the table type to be created and, optionally, in which schema the creation should take place.

<column_definition> ::= <column_name> <data_type> [<column_store_data_type>] [<ddic_data_type>]




<data_type> ::= DATE | TIME | SECONDDATE | TIMESTAMP | TINYINT | SMALLINT | INTEGER | BIGINT | SMALLDECIMAL | DECIMAL | REAL | DOUBLE | VARCHAR | NVARCHAR | ALPHANUM | SHORTTEXT | VARBINARY | BLOB | CLOB | NCLOB | TEXT


178 2012-12-19


<ddic_data_type> ::= DDIC_ACCP | DDIC_ALNM | DDIC_CHAR | DDIC_CDAY | DDIC_CLNT | DDIC_CUKY | DDIC_CURR | DDIC_D16D | DDIC_D34D | DDIC_D16R | DDIC_D34R | DDIC_D16S | DDIC_D34S | DDIC_DATS | DDIC_DAY | DDIC_DEC | DDIC_FLTP | DDIC_GUID | DDIC_INT1 | DDIC_INT2 | DDIC_INT4 | DDIC_INT8 | DDIC_LANG | DDIC_LCHR | DDIC_MIN | DDIC_MON | DDIC_LRAW | DDIC_NUMC | DDIC_PREC | DDIC_QUAN | DDIC_RAW | DDIC_RSTR | DDIC_SEC | DDIC_SRST | DDIC_SSTR | DDIC_STRG | DDIC_STXT | DDIC_TIMS | DDIC_UNIT | DDIC_UTCM | DDIC_UTCL | DDIC_UTCS | DDIC_TEXT | DDIC_VARC | DDIC_WEEK

The available data types. Please see Data Types

Description

The CREATE TYPE statement creates a user-defined type.The syntax for defining table types follows the SQL syntax for defining new types. The table type is specifiedusing a list of attribute names and primitive data types. For each table type, attributes must have uniquenames.

Example

You create a table type called tt_publishers.

CREATE TYPE tt_publishers AS TABLE ( publisher INTEGER, name VARCHAR(50), price DECIMAL, cnt INTEGER);

You create a table type called tt_years.

CREATE TYPE tt_years AS TABLE ( year VARCHAR(4), price DECIMAL, cnt INTEGER);

DROP FUNCTION

Syntax

DROP FUNCTION <func_name> [<drop_option>]

Syntax Elements

<func_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The name of the function to be dropped, with optional schema name.


When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedfunction, dependent objects of the function will be invalidated but not dropped.


CASCADE

Drops the function and dependent objects.

RESTRICT


179 2012-12-19

Drops the function only when dependent objects do not exist. If this drop option is used and a dependentobject exists an error will be thrown.

Description

Drops a function created using CREATE FUNCTION from the database catalog.

Examples

You drop a function called my_func from the database using a non-cascaded drop.

DROP FUNCTION my_func;

DROP PROCEDURE

Syntax

DROP PROCEDURE <proc_name> [<drop_option>]

Syntax Elements


The name of the procedure to be dropped, with optional schema name.


When <drop_option> is not specified a non-cascaded drop will be performed. This will only drop the specifiedprocedure, dependent objects of the procedure will be invalidated but not dropped.


CASCADE

Drops the procedure and dependent objects.

RESTRICT

Drops the procedure only when dependent objects do not exist. If this drop option is used and a dependentobject exists an error will be thrown.

Description

Drops a procedure created using CREATE PROCEDURE from the database catalog.

Examples

You drop a procedure called my_proc from the database using a non-cascaded drop.

DROP PROCEDURE my_proc;

DROP TYPE

Syntax

DROP TYPE <type_name> [<drop_option>]

Syntax Elements

<type_name> ::= [<schema_name>.]<identifier> <schema_name> ::= <identifier>

The identifier of the table type to be droppped, with optional schema name.



180 2012-12-19

When <drop_option> is not specified a non-cascaded drop will be performed. This will drop only the specifiedtype, dependent objects of the type will be invalidated but not dropped.


Description

The DROP TYPE statement removes a user-defined table type.

Example

You create a table type called my_type.

CREATE TYPE my_type AS TABLE ( column_a DOUBLE );

You drop the my_type table type.

DROP TYPE my_type;


181 2012-12-19

Restrictions for SQL Statements

The table below shows the maximum allowable limit for each entry.


182 2012-12-19

Database

Database size limit Limited by storage size RS: 1TB

Number of locks Unlimited for record locks,16384 for table locks

Number of sessions 8192

Schemas

Number of tables in a schema 131072

Identifier length 127 characters

Length of an alias name 128 characters

Table name length Please see "Identifier length" above

Column name length Please see "Identifier length" above

Length of a constant string literal 32767 bytes

Number of hex characters in a binary literal 8192

Tables and Views

Number of columns in a table 1000

Number of columns in a view 1000

Number of partitions of a column table 1000

Number of rows in each table

Limited by storage size RS:1TB/sizeof(row),

CS: 2^31 * number of partitions

Length of a row Limited by RS storage size ( 1TB )

Size of a non-partitioned table Limited by RS storage size ( 1TB )

Indexes and Constraints

Number of indexes for each table 1023

Number of primary key columns in each table 16

Number of columns in an index 16

Number of columns in a UNIQUE constraint 16

Size of sum of primary key, index, UNIQUE

constraint16384

SQL


183 2012-12-19

Length of an SQL statement 2GB

Depth of SQL view nesting 128

Depth of SQL parse tree 255

Number of joined tables in an SQL statement or view 255

Number of columns in an ORDER BY, GROUP BY or SELECTclause

65535

Number of elements in predicates 65535

Number of elements in SELECT clause 65535

SQLScript

Size of all stored procedures Limited by RS s torage s i ze ( 1TB )

Note:You can obtain the current system limits from your SAP HANA Database instance by using

select * from m_system_limits

See Also

M_SYSTEM_LIMITS in the System Tables and Monitor Views Reference


184 2012-12-19

SQL Error Codes

The following table lists error codes displayed by SAP HANA database and their descriptions.


185 2012-12-19

ErrorCode

Description

1 Genera l warning

2 Genera l error

3 Fata l error

4 Cannot a l locate enough memory

5 Ini tia l i zation error

6 Inva l id data

7 Feature not supported

8 Inva l id argument

9 Index out of bounds

10 Inva l id username or password

11 Inva l id s tate

12 Cannot open fi le

13 Cannot create/wri te fi le

14 Cannot a l locate enough disk space

15 Cannot find fi le

16 Statement retry

17 Metadata schema vers ion incompatible between database and executable fi le

18 Service shutting down

19 Inva l id l i cense

128 Trans action error

129 Trans action rol led back by an interna l error

130 Trans action rol led back by integri ty constra int violation

131 Trans action rol led back by lock wai t timeout

132 Trans action rol led back due to unava i lable resource

133 Trans action rol led back by detected deadlock

134 Fa i lure in access ing checkpoint fi le

135 Fa i lure in access ing anchor fi le

136 Fa i lure in access ing log fi le

137 Fa i lure in access ing archive fi le

138 Trans action seria l i zation fa i lure

139 Current operation cancel led by request and transaction rol led back

140 Inva l id wri te-transaction identi fier

141 Fa i lure in access ing invis ible log fi le

142 Exceed max num of concurrent transactions

143 Trans action seria l i zation fa i lure unti l timeout expi res

144 Trans action rol lback, unique constra int violated

145 Trans action dis tribution work fa i lure

146 Res ource busy and acquire with NOWAIT speci fied

147 Incons is tency between data and log

148 Trans action s tart i s blocked unti l Master_Restart fini shes

149 Dis tributed transaction commit fa i lure

150 Statement cancel led due to old snapshot

256 SQL process ing error

257 SQL syntax error

258 Ins ufficient privi lege

259 Inva l id table name

260 Inva l id column name

261 Inva l id index name

262 Inva l id query name

263 Inva l id a l ias name

264 Inva l id datatype

265 Expres s ion miss ing

266 Incons is tent datatype

267 Speci fied length too long for i ts datatype

268 Column ambiguous ly defined


186 2012-12-19

269 Too many va lues

270 Not enough va lues

271 Dupl icate a l ias

272 Dupl icate column name

273 Not a s ingle character s tring

274 Ins erted va lue too large for column

275 Aggregate function not a l lowed

276 Mis s ing aggregation or grouping

277 Not a GROUP BY express ion

278 Nested group function without GROUP BY

279 Group function i s nested

280 ORDER BY i tem must be the number of a SELECT-l i s t

281 Outer join not a l lowed in operand of OR or IN

282 Two tables cannot be outer-joined to each other

283 A table may be outer joined to at most one other table

284 Join field does not match

285 Inva l id join condi tion

286 Identi fier i s too long

287 Cannot insert NULL or update to NULL

288 Cannot use dupl icate table name

289 Cannot use dupl icate index name

290 Cannot use dupl icate query name

291 Argument identi fier must be pos i tive

292 wrong number of arguments

293 Argument type mismatch

294 Cannot have more than one primary key

295 Too long multi key length

296 Repl icated table must have a primary key

297 Cannot update primary key field in repl icated table

298 Cannot s tore DDL

299 Cannot drop index used for enforcement of unique/primary key

300 Argument index i s out of range

301 Unique constra int violated

302 Inva l id CHAR or VARCHAR va lue

303 Inva l id DATE, TIME or TIMESTAMP va lue

304 Divis ion by zero undefined

305 Single-row query returns more than one row

306 Inva l id cursor

307 Numeric va lue out of range

308 Column name a l ready exis ts

309 Correlated subquery cannot have TOP or ORDER BY

310 SQL error in procedure

311 Cannot drop a l l columns in a table

312 Sequence i s exhausted

313 Inva l id sequence

314 Numeric overflow

315 Inva l id synonym

316 wrong number of arguments in function invocation

317 P_QUERYPLANS not exis ts nor va l id format

318 Decimal precis ion speci fier i s out of range

319 Decimal sca le speci fier i s out of range

320 Cannot create index on express ion with datatype LOB

321 Inva l id view name

322 Cannot use dupl icate view name

323 Dupl icate repl ication ID

324 Cannot use dupl icate sequence name

325 Inva l id escape sequence


187 2012-12-19

326 CURRVAL of given sequence i s not yet defined in this sess ion

327 Cannot expla in plan of given s tatement

328 Inva l id name of function or procedure

329 Cannot use dupl icate name of function or procedure

330 Cannot use dupl icate synonym name

331 User name a l ready exis ts

332 Inva l id user name

333 Column not a l lowed

334 Inva l id user privi lege

335 Field a l ias name a l ready exis ts

336 Inva l id defaul t va lue

337 INTO clause not a l lowed for this SELECT s tatement

338 Zero-length col lumns are not a l lowed

339 Inva l id number

340 Not a l l variables bound

341 Numeric underflow

342 Col la tion confl i ct

343 Inva l id col la te name

344 Parse error in data loader

345 Not a repl ication table

346 Inva l id repl ication ID

347 Inva l id option in monitor

348 Inva l id datetime format

349 Cannot CREATE UNIQUE INDEX

350 Cannot drop columns in the primary-key column l i s t

351 Column i s referenced in a multi -column constra int

352 Cannot create unique index on CDX table

353 Update log group name a l ready exis ts

354 Inva l id update log group name

355 The base table of the update log table must have a primary key

356 Exceed maximum number of update log group

357 The base table a l ready has a update log table

358 Update log table can not have a update log table

359 Concatenated s tring i s too long

360 View WITH CHECK OPTION where-clause violation

361 Data manipulation operation not lega l on this view

362 Inva l id schema name

363 Number of index columns exceeds i ts maximum

364 Inva l id partia l key s i ze

365 No matching unique or primary key for this column l i s t

366 Referenced table does not have a primary key

367 Number of referencing columns must match referenced columns

368 Unique constra int not a l lowed on temporary table

369 Exceed maximum view depth l imit

370 Cannot perform DIRECT INSERT operation on table with unique indexes

371 Inva l id XML document

372 Inva l id XPATH

373 Inva l id XML duration va lue

374 Inva l id XML function usage

375 Inva l id XML index operation

376 Python s tored procedure error

377 JIT operation error

378 Inva l id column view

379 Table s chema mismatch

380 Fa i l to change run level

381 Fa i l to res tart

382 Fa i l to col lect a l l vers ion garbage


188 2012-12-19

383 Inva l id identi fier

384 Constant s tring i s too long

385 Could not res tore sess ion

386 Cannot use dupl icate schema name

387 Table ambiguous ly defined

388 Role a l ready exis ts

389 Inva l id role name

390 Inva l id user type

391 Inva l idated view

392 Can't ass ign cycl i c role

393 Roles must not receive a privi lege with grant option

394 Error revoking role

395 Inva l id user-defined type name

396 Cannot use dupl icate user-defined type name

397 Inva l id object name

398 Cannot have more than one order by

399 Role tree too deep

400 Primary key not a l lowed on insert-only table

401 Unique constra int not a l lowed on insert-only table

402 The user was a l ready dropped before query execution

403 Interna l error

404 Inva l id (non-exis tent) s tructured privi lege name

405 Cannot use dupl icate s tructured privi lege name

406 INSERT, UPDATE and UPSERT are disa l lowed on the generated field

407 Inva l id date format

408 Pass word or parameter required for user

409 Multiple va lues for a parameter not supported

410 Inva l id privi lege namespace

411 Inva l id table type

412 Inva l id password layout

413 Last n passwords can not be reused

414 User i s forced to change password

415 User i s deactivated

416 User i s locked

417 Can't drop without CASCADE speci fi cation

418 Inva l id view query for creation

419 Can't drop with RESTRICT speci fi cation.

420 Pass word change currently not a l lowed

421 Cannot create ful l text index

422 Privi leges must be ei ther a l l SQL or a l l from one namespace

423 LiveCache error

424 Inva l id name of package

425 Dupl icate package name

426 Number of columns mismatch

427 Cannot reserve index ID any more

428 Inva l id query plan ID

429 Integri ty check fa i led

430 Inva l idated procedure

431 User's password wi l l expi re within few days

432 This s yntax has been deprecated and wi l l be removed in next release

433 Nul l va lue found

434 Inva l id object ID

435 Inva l id express ion

436 Could not set sys tem l i cense

437 Only commands for l i cense handl ing are a l lowed in current s tate

438 Inva l id user parameter va lue

439 Compos i te error


189 2012-12-19

440 Table type convers ion error

441 This feature has been deprecated and wi l l be removed in next release

442 Number of columns exceeds i ts maximum

443 Inva l id ca lculation scenario name

444 Package manager error

512 Repl ication error

513 Cannot execute DDL s tatement on repl ication table whi le repl icating

514 Fa i lure in access ing anchor fi le

515 Fa i lure in access ing log fi le

516 Repl ication table has not confl i ct report table

517 Confl i ct report table a l ready enabled

518 Confl i ct report table a l ready disabled

576 API error

577 Cursor type of forward i s not a l lowed

578 Inva l id s tatement

579 Exceed maximum batch s i ze

580 Server rejected the connection(protocol vers ion mismatch)

581 This function can be ca l led only in the case of s ingle s tatement

582 This query does not have resul t set

583 Connection does not exis t

584 No more lob data

585 Operation i s not permitted

586 Inva l id parameter i s received from server

587 Res ul t set i s currently inva l id

588 Next() i s not ca l led for this resul t set

589 Too many parameters are set

590 Some paramters are miss ing

591 Interna l error

592 Not supported type convers ion

593 Remote-only function

594 No more resul t row in resul t set

595 Speci fied parameter i s not output parameter

596 LOB s treaming i s not permitted in auto-commit mode

597 Sess ion context error

598 Fa i led to execute the externa l s tatement

599 Sess ion layer i s not ini tia l i zed yet

600 Fa i led routed execution

601 Too many sess ion variables are set

602 Cannot set readonly sess ion variable

603 Inva l id LOB

604 Remote temp table access fa i lure

605 Inva l id XA join request

606 Exceed maximum LOB s i ze

607 Fa i led to cleanup resources

608 Exceed maximum number of prepared s tatements

1024 Sess ion error

1025 Communication error

1026 Cannot bind a communication port

1027 Communication ini tia l i zation error

1028 I/O control error

1029 Connection fa i lure

1030 Send error

1031 Receive error

1032 Cannot create a thread

1033 Error whi le pars ing protocol

1034 Exceed maximum number of sess ions

1035 Not supported vers ion


190 2012-12-19

1036 Inva l id sess ion ID

1037 Unknown hostname

1280 SqlScript error

1281 Wrong number or types of parameters in ca l l

1282 Output parameter not avariable

1283 OUT and IN OUT parameters may not have defaul t expres s ions

1284 Dupl icate parameters are not permitted

1285 At mos t one declaration i s permitted in the declaration section

1286 Cursor must be declared by SELECT s tatement

1287 Identi fier must be declared

1288 Expres s ion cannot be used as an ass ignment target

1289 Expres s ion cannot be used as an INTO-target of SELECT/FETCH s tatement

1290 Expres s ion i s inappropriate as the left hand s ide of an ass ignment s tatement

1291 Expres s ion i s of wrong type

1292 Il lega l EXIT s tatement, i t must appear ins ide a loop

1293 Identi fier name must be an exception name

1294 An INTO clause i s expected in SELECT s tatement

1295 EXPLAIN PLAN and CALL s tatement are not a l lowed

1296 Identi fier i s not a cursor

1297 Wrong number of va lues in the INTO l i s t of a FETCH s tatement

1298 Unhandled user-defined exception

1299 No data found

1300 Fetch returns more than requested number of rows

1301 Numeric or va lue error

1302 Para l lel i zable function cannot have OUT or IN OUT parameter

1303 User-defined exception

1304 Cursor i s a l ready opened

1305 Return type i s inva l id

1306 Return type mismatch

1307 Unsupported datatype i s used

1308 Il lega l s ingle ass ignment

1309 Inva l id use of table variable

1310 Sca lar type i s not a l lowed

1311 Out parameter i s not speci fied

1312 At mos t one output parameter i s a l lowed

1313 Output parameter should be a table or a table variable

1314 Inappropriate variable name: do not a l low "" for the name of variable or parameter

1315 Return resul t set from select s tmt exis t when resul t view i s defined

1316 Some out table var i s not ass igned

1317 Function name exceedes max. l imit

1318 Bui l t-in function not defined

1319 Parameter must be a table name

1320 Parameter must be an attribute name without a table name upfront

1321 Parameter must be an attribute name without an a l ias

1322 CE_CALC not a l lowed

1323 Parameter must be a vector of columns or aggregations

1324 Join attribute must be ava i lable in projection l i s t

1325 Parameter must be a vector of SQL identi fiers

1326 Dupl icate attribute name

1327 Parameter has a non supported type

1328 Attribute not found in column table

1329 Dupl icate column name

1330 Syntax Error for ca lculated Attribute

1331 Syntax Error in fi l ter express ion

1332 Parameter must be a va l id column table name

1333 Join attributes not found in variable

1334 Input parameters do not have the same table type


191 2012-12-19

1335 Cycl ic dependency found in a runtime procedure

1336 Unexpected interna l exception caught in a runtime procedure

1337 Variable depends on an unass igned variable

1338 CE_CONVERSION: customizing table miss ing

1339 Too many parameters

1340 The depth of the nested ca l l i s too deep

1536 Swapx error

1537 This table has no swap space

1538 Swap a l ready activated

1539 Swap not yet activated

1540 Swap s pace i s not created

1541 Fa i lure in unpinning a swap page

1542 Fa i lure in swap fi le

1543 Fa i lure in access ing swap data fi le

1544 Fa i lure in access ing swap log fi le

1545 Swap buffer overflow

1546 Swap buffer reservation fa i lure

1792 Shared memory error

1793 Inva l id key or inva l id s i ze

1794 The segment a l ready exis ts

1795 Exceed the sys tem-wide l imit on shared memory

1796 No segment exis ts for the given key, and IPC_CREAT was not speci fied

1797 The user does not have permiss ion to access the shared memory s egment

1798 No memory could be a l located for segment overhead

1799 Inva l id shmid

1800 Al low read access for shmid

1801 Shmid points to a removed identi fier

1802 The effective user ID of the ca l l ing process i s not the creator

1803 The GID or UID va lue i s too large to be s tored in the s tructure

1804 The user does not have permiss ion to access the shared memory s egment

1805 Inva l id shmid

1806 No memory could be a l located for the descriptor or for the page tables

1807 Unknown shared memory error

2048 Column s tore error

2049 Primary key i s not speci fied for column table

2050 Not supported ddl type for column table

2051 Not supported data type for column table

2052 Not supported dml type for column table

2053 Inva l id returned va lue from attribute engine

2304 Python DBAPI error

2305 Interface fa i lure

2306 Programming mis take

2307 Inva l id query parameter

2308 Not supported encoding for s tring

2560 Dis tributed metadata error

2561 DDL redi rect error

2562 DDL noti fi cation error

2563 DDL inva l id conta iner ID

2564 DDL inva l id index ID

2565 Dis tributed envi ronment error

2566 Network error

2567 Metadata update not supported in s lave

2568 Metadata update of master indexserver i s fa i led

2816 SqlScript Error

2817 SqlScript Bui l tin Function

2818 -2889

SqlScript

3584 Dis tributed SQL error


192 2012-12-19

3585 Expres s ion shipping fa i lure

3586 Operator shipping fa i lure

3587 Inva l id protocol or indexserver (s tati s ti csserver) shutdown during dis tributed query execution

3588 Sequence shipping fa i lure

3589 Remote query exectuion fa i lure

3840 genera l audi ting error

3841 Inva l id privi lege

3842 Audit tra i l wri ter i s blocked

3843 Audit pol icy wi th current name a l ready exis ts

3844 Inva l id combination of audi t actions

3845 Inva l id action s tatus for audi ting

3846 Inva l id audi ting level

3847 Inva l id pol icy name

4096 Genera l error in the process of s tored plan

4097 Inva l id operation in generating plans

4098 Inva l id operation in execution the chosen plan

4099 Inva l id operation in s toring the pinned plan

4100 Inva l id operation in loading the s tored plan

4101 Inva l id operation in deleting the chosen plan

4103 Fa i led to prepare for runtime reorganization

4104 Trans action blocked s ince runtime reorganization i s in progres s

4105ERR_REORG_TRANS_EXISTS_GENERAL. Cannot s tart reorganization due to the transactions inexecution


193 2012-12-19

hana_sql_ref_en

Documents