Chp 6 Physical Database Design and Performance

© 2009 Pearson Education, Inc. Publishing as © 2009 Pearson Education, Inc. Publishing as Prentice HallPrentice Hall 11

Chapter 6:Chapter 6:Physical Database Design Physical Database Design

and Performanceand Performance

Modern Database Modern Database ManagementManagement99thth Edition Edition

Jeffrey A. Hoffer, Mary B. Prescott, Jeffrey A. Hoffer, Mary B. Prescott, Heikki TopiHeikki Topi

Chapter 6 © 2009 Pearson Education, Inc. Publishing as Prentice Hall© 2009 Pearson Education, Inc. Publishing as Prentice Hall 22

ObjectivesObjectives Definition of termsDefinition of terms Describe the physical database design Describe the physical database design

processprocess Choose storage formats for attributesChoose storage formats for attributes Select appropriate file organizationsSelect appropriate file organizations Describe three types of file organizationDescribe three types of file organization Describe indexes and their appropriate useDescribe indexes and their appropriate use Translate a database model into efficient Translate a database model into efficient

structuresstructures Know when and how to use denormalizationKnow when and how to use denormalization


Physical Database DesignPhysical Database Design

Purpose–translate the logical Purpose–translate the logical description of data into the description of data into the technical technical specificationsspecifications for storing and for storing and retrieving dataretrieving data

Goal–create a design for storing data Goal–create a design for storing data that will provide that will provide adequate adequate performanceperformance and insure and insure database database integrityintegrity, , securitysecurity, and , and recoverabilityrecoverability


Physical Design ProcessPhysical Design Process

Normalized relations

Volume estimates

Attribute definitions

Response time expectations

Data security needs

Backup/recovery needs

Integrity expectations

DBMS technology used

Inputs

Attribute data types

Physical record descriptions (doesn’t always match logical design)

File organizations

Indexes and database architectures

Query optimization

Leads to

Decisions


Figure 6-1 Composite usage map (Pine Valley Furniture Company)


Figure 6-1 Composite usage map (Pine Valley Furniture Company) (cont.)

Data volumes



Access Frequencies (per hour)



Usage analysis:140 purchased parts accessed per hour 80 quotations accessed from these 140 purchased part accesses 70 suppliers accessed from these 80 quotation accesses



Usage analysis:75 suppliers accessed per hour 40 quotations accessed from these 75 supplier accesses 40 purchased parts accessed from these 40 quotation accesses


Designing FieldsDesigning Fields

Field: smallest unit of data in Field: smallest unit of data in databasedatabase

Field design Field design Choosing data typeChoosing data type Coding, compression, encryptionCoding, compression, encryption Controlling data integrityControlling data integrity


Choosing Data TypesChoosing Data Types

CHAR–fixed-length characterCHAR–fixed-length character VARCHAR2–variable-length character VARCHAR2–variable-length character

(memo)(memo) LONG–large numberLONG–large number NUMBER–positive/negative numberNUMBER–positive/negative number INEGER–positive/negative whole numberINEGER–positive/negative whole number DATE–actual dateDATE–actual date BLOB–binary large object (good for graphics, BLOB–binary large object (good for graphics,

sound clips, etc.)sound clips, etc.)


Figure 6-2 Example code look-up table(Pine Valley Furniture Company)

Code saves space, but costs an additional lookup to obtain actual value


Field Data IntegrityField Data Integrity Default value–assumed value if no Default value–assumed value if no

explicit valueexplicit value Range control–allowable value limitations Range control–allowable value limitations

(constraints or validation rules)(constraints or validation rules) Null value control–allowing or prohibiting Null value control–allowing or prohibiting

empty fieldsempty fields Referential integrity–range control (and Referential integrity–range control (and

null value allowances) for foreign-key to null value allowances) for foreign-key to primary-key match-upsprimary-key match-ups

Sarbanes-Oxley Act (SOX) legislates importance of financial data integrity


Handling Missing DataHandling Missing Data

Substitute an estimate of the missing Substitute an estimate of the missing value (e.g., using a formula)value (e.g., using a formula)

Construct a report listing missing valuesConstruct a report listing missing values In programs, ignore missing data unless In programs, ignore missing data unless

the value is significant (sensitivity the value is significant (sensitivity testing)testing)

Triggers can be used to perform these operations


Physical RecordsPhysical Records Physical Record: A group of fields Physical Record: A group of fields

stored in adjacent memory locations stored in adjacent memory locations and retrieved together as a unitand retrieved together as a unit

Page: The amount of data read or Page: The amount of data read or written in one I/O operationwritten in one I/O operation

Blocking Factor: The number of Blocking Factor: The number of physical records per pagephysical records per page


DenormalizationDenormalization Transforming Transforming normalizednormalized relations into relations into

unnormalizedunnormalized physical record specifications physical record specifications Benefits:Benefits:

Can improve performance (speed) by reducing number of Can improve performance (speed) by reducing number of table lookups (i.e. table lookups (i.e. reduce number of necessary join queriesreduce number of necessary join queries))

Costs (due to data duplication)Costs (due to data duplication) Wasted storage spaceWasted storage space Data integrity/consistency threatsData integrity/consistency threats

Common denormalization opportunitiesCommon denormalization opportunities One-to-one relationship (Fig. 6-3)One-to-one relationship (Fig. 6-3) Many-to-many relationship with attributes (Fig. 6-4)Many-to-many relationship with attributes (Fig. 6-4) Reference data (1:N relationship where 1-side has data not Reference data (1:N relationship where 1-side has data not

used in any other relationship) (Fig. 6-5)used in any other relationship) (Fig. 6-5)


Figure 6-3 A possible denormalization situation: two entities with one-to-one relationship


Figure 6-4 A possible denormalization situation: a many-to-many relationship with nonkey attributes

Extra table access required

Null description possible


Figure 6-5A possible denormalization situation:reference data

Extra table access required

Data duplication


PartitioningPartitioning Horizontal Partitioning: Distributing the rows of Horizontal Partitioning: Distributing the rows of

a table into several separate filesa table into several separate files Useful for situations where different users need Useful for situations where different users need

access to different rowsaccess to different rows Three types: Key Range Partitioning, Hash Three types: Key Range Partitioning, Hash

Partitioning, or Composite PartitioningPartitioning, or Composite Partitioning Vertical Partitioning: Distributing the columns Vertical Partitioning: Distributing the columns

of a table into several separate relationsof a table into several separate relations Useful for situations where different users need Useful for situations where different users need

access to different columnsaccess to different columns The primary key must be repeated in each fileThe primary key must be repeated in each file

Combinations of Horizontal and VerticalCombinations of Horizontal and Vertical

Partitions often correspond with User Schemas (user views)


Partitioning (cont.)Partitioning (cont.) Advantages of Partitioning:Advantages of Partitioning:

Efficiency: Records used together are grouped togetherEfficiency: Records used together are grouped together Local optimization: Each partition can be optimized for Local optimization: Each partition can be optimized for

performanceperformance Security, recoverySecurity, recovery Load balancing: Partitions stored on different disks, Load balancing: Partitions stored on different disks,

reduces contentionreduces contention Take advantage of parallel processing capabilityTake advantage of parallel processing capability

Disadvantages of Partitioning:Disadvantages of Partitioning: Inconsistent access speed: Slow retrievals across Inconsistent access speed: Slow retrievals across

partitionspartitions Complexity: Non-transparent partitioningComplexity: Non-transparent partitioning Extra space or update time: Duplicate data; access from Extra space or update time: Duplicate data; access from

multiple partitionsmultiple partitions


Oracle 10i Horizontal Oracle 10i Horizontal Partitioning MethodsPartitioning Methods

Key range partitioningKey range partitioning Partitions defined by range of field valuesPartitions defined by range of field values Could result in unbalanced distribution of rowsCould result in unbalanced distribution of rows Like-valued fields share partitionsLike-valued fields share partitions

Hash partitioningHash partitioning Partitions defined via hash functionsPartitions defined via hash functions Will guarantee balanced distribution of rowsWill guarantee balanced distribution of rows Partition could contain widely varying valued Partition could contain widely varying valued

fieldsfields Composite partitioningComposite partitioning

Combines key range and hashCombines key range and hash


Data ReplicationData Replication

Purposely storing the same data in Purposely storing the same data in multiple locations of the databasemultiple locations of the database

Improves performance by allowing Improves performance by allowing multiple users to access the same data multiple users to access the same data at the same time with minimum at the same time with minimum contentioncontention

Sacrifices data integrity due to data Sacrifices data integrity due to data duplicationduplication

Best for data that is not updated oftenBest for data that is not updated often


Designing Physical FilesDesigning Physical Files Physical File: Physical File:

A named portion of secondary memory allocated A named portion of secondary memory allocated for the purpose of storing physical recordsfor the purpose of storing physical records

Tablespace–named set of disk storage elements Tablespace–named set of disk storage elements in which physical files for database tables can be in which physical files for database tables can be storedstored

Extent–contiguous section of disk spaceExtent–contiguous section of disk space Constructs to link two pieces of data:Constructs to link two pieces of data:

Sequential storageSequential storage Pointers–field of data that can be used to locate Pointers–field of data that can be used to locate

related fields or recordsrelated fields or records


Figure 6-6 Physical file terminology in an Oracle environment


File OrganizationsFile Organizations Technique for physically arranging records of a file Technique for physically arranging records of a file

on secondary storageon secondary storage Factors for selecting file organization:Factors for selecting file organization:

Fast data retrieval and throughputFast data retrieval and throughput Efficient storage space utilizationEfficient storage space utilization Protection from failure and data lossProtection from failure and data loss Minimizing need for reorganizationMinimizing need for reorganization Accommodating growthAccommodating growth Security from unauthorized useSecurity from unauthorized use

Types of file organizationsTypes of file organizations SequentialSequential IndexedIndexed HashedHashed


Figure 6-7a Sequential file organization

If not sortedAverage time to find desired record = n/2

1

2

n

Records of the file are stored in sequence by the primary key field values

If sorted – every insert or delete requires resort


Indexed File OrganizationsIndexed File Organizations Index–a separate table that contains Index–a separate table that contains

organization of records for quick retrievalorganization of records for quick retrieval Primary keys are automatically indexedPrimary keys are automatically indexed Oracle has a CREATE INDEX operation, and Oracle has a CREATE INDEX operation, and

MS ACCESS allows indexes to be created for MS ACCESS allows indexes to be created for most field typesmost field types

Indexing approaches:Indexing approaches: B-tree index, Fig. 6-7bB-tree index, Fig. 6-7b Bitmap index, Fig. 6-8Bitmap index, Fig. 6-8 Hash Index, Fig. 6-7cHash Index, Fig. 6-7c Join Index, Fig 6-9Join Index, Fig 6-9


Figure 6-7b B-tree index

uses a tree searchAverage time to find desired record = depth of the tree

Leaves of the tree are all at same level

consistent access time


Figure 6-7cHashed file or index organization

Hash algorithmUsually uses division-remainder to determine record position. Records with same position are grouped in lists


Figure 6-8Bitmap index organization

Bitmap saves on space requirementsRows - possible values of the attribute

Columns - table rows

Bit indicates whether the attribute of a row has the values


Figure 6-9 Join Indexes–speeds up join operations



Clustering FilesClustering Files

In some relational DBMSs, related records In some relational DBMSs, related records from different tables can be stored from different tables can be stored together in the same disk areatogether in the same disk area

Useful for improving performance of join Useful for improving performance of join operationsoperations

Primary key records of the main table are Primary key records of the main table are stored adjacent to associated foreign key stored adjacent to associated foreign key records of the dependent tablerecords of the dependent table

e.g. Oracle has a CREATE CLUSTER e.g. Oracle has a CREATE CLUSTER commandcommand


Rules for Using IndexesRules for Using Indexes

1.1. Use on larger tablesUse on larger tables2.2. Index the primary key of each tableIndex the primary key of each table3.3. Index search fields (fields frequently Index search fields (fields frequently

in WHERE clause)in WHERE clause)4.4. Fields in SQL ORDER BY and GROUP Fields in SQL ORDER BY and GROUP

BY commandsBY commands5.5. When there are >100 values but When there are >100 values but

not when there are <30 valuesnot when there are <30 values


Rules for Using Indexes Rules for Using Indexes (cont.)(cont.)

6.6. Avoid use of indexes for fields with long Avoid use of indexes for fields with long values; perhaps compress values firstvalues; perhaps compress values first

7.7. DBMS may have limit on number of indexes DBMS may have limit on number of indexes per table and number of bytes per indexed per table and number of bytes per indexed field(s)field(s)

8.8. Null values will not be referenced from an Null values will not be referenced from an indexindex

9.9. Use indexes heavily for non-volatile Use indexes heavily for non-volatile databases; limit the use of indexes for databases; limit the use of indexes for volatile databasesvolatile databases

Why? Because modifications (e.g. inserts, deletes) require Why? Because modifications (e.g. inserts, deletes) require updates to occur in index filesupdates to occur in index files


RAIDRAID

Redundant Array of Inexpensive Redundant Array of Inexpensive DisksDisks

A set of disk drives that appear to A set of disk drives that appear to the user to be a single disk drivethe user to be a single disk drive

Allows parallel access to data Allows parallel access to data (improves access speed)(improves access speed)

Pages are arranged in Pages are arranged in stripesstripes


Figure 6-10RAID with four disks and striping

Here, pages 1-4 can be read/written simultaneously


Raid Types (Figure 6-10)Raid Types (Figure 6-10) Raid 0

Maximized parallelism No redundancy No error correction no fault-tolerance

Raid 1 Redundant data–fault tolerant Most common form

Raid 2 No redundancy One record spans across data

disks Error correction in multiple

disks–reconstruct damaged data

Raid 3 Error correction in one disk Record spans multiple data

disks (more than RAID2) Not good for multi-user

environments, Raid 4

Error correction in one disk Multiple records per stripe Parallelism, but slow updates

due to error correction contention

Raid 5 Rotating parity array Error correction takes place in

same disks as data storage Parallelism, better performance

than Raid4


Data

base

D

ata

base

A

rch

itect

ure

s

Arc

hit

ect

ure

s (

Figure

(F

igure

6

-11

)6

-11

)

Legacy

Systems

Current

Technology

Data

Warehouses


Query OptimizationQuery Optimization Parallel query processing – possible when Parallel query processing – possible when

working in multiprocessor systemsworking in multiprocessor systems Overriding automatic query optimization – Overriding automatic query optimization –

allows for query writers to preempt the allows for query writers to preempt the automated optimizationautomated optimization

Picking data block size – factors to Picking data block size – factors to consider include:consider include: Block contention, random and sequential row Block contention, random and sequential row

access speed, row sizeaccess speed, row size Balancing I/O across disk controllersBalancing I/O across disk controllers


Query Design GuidelinesQuery Design Guidelines Understand how indexes are usedUnderstand how indexes are used Keep optimization statistics up-to-dateKeep optimization statistics up-to-date Use compatible data types for fields and Use compatible data types for fields and

literalsliterals Write simple queriesWrite simple queries Break complex queries into multiple, Break complex queries into multiple,

simple partssimple parts Don’t use one query inside anotherDon’t use one query inside another Don’t combine a table with itselfDon’t combine a table with itself


Query Design Guidelines Query Design Guidelines (cont.)(cont.)

Create temporary tables for groups of Create temporary tables for groups of queriesqueries

Combine update operationsCombine update operations Retrieve only the data you needRetrieve only the data you need Don’t have the DBMS sort without an indexDon’t have the DBMS sort without an index Learn!Learn! Consider the total query processing time Consider the total query processing time

for ad hoc queriesfor ad hoc queries


All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic,

mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Printed in the United States of America.

Copyright © 2009 Pearson Education, Inc. Publishing as Prentice Copyright © 2009 Pearson Education, Inc. Publishing as Prentice HallHall

Chp 6 Physical Database Design and Performance

Documents

physical database design

key range partitioning

file organizations

primary key

denormalization situation

physical records

tree index

reference data