Clusterpoint Margarita Sudņika ms11077
Jan 19, 2016
Clusterpoint
Margarita Sudņikams11077
RDBMS & NoSQL
Databases & tables → Document stores
Columns, rows → Schemaless documents
Scales UP → Scales UP & OUT
Replications → Sharding & Replications
For table like data → Unstructured data
Legacy & mature → New
21.04.23
Clusterpoint
A scalable high-speed NoSQL database technology with Google-like search
Manually ranking (svara piešķiršana)
Solves 2 big data access problems: Long time waiting for query execution
Querry execution 0,005-0,5 seconds Loads of information
21.04.23
Arhitecture
21.04.23
Application
Clusterpoint server
software
HT
TP
DOCUMENTS
MAILS
CONTACTS
PROJECTS
CUSTOMERS
EMPLOYEES
CLUSTER NODES (multi-server
hardware)
STAND-ALONE SERVER
Clusterpoint
Data storage model xml
Supported formats Json Xml HTML Text
21.04.23
Features
Full context search
Unlimted database size
Guaranteed querry size <0,5 s
Clustering as default feature
Scallable database mirroring
Snippets with search hits
Web friendly api
Flexible data relevancy rules21.04.23
21.04.23
Access
21.04.23
SearchFree text
Phrase
Wildcards
Patterns matches by lookup John Smith
In XML database structure
Did you mean “...?” feature
Faceted search and navigation
Full data index for xml data21.04.23
API
Simple, robust XML messagingXML request/response similar to SOAP
Transport– http, https (post, get)– tcp– unix domain socket
> 20 API commands
Libraries: PHP, .NET (web service)
21.04.23
API message
21.04.23
<?xml version=”1.0” encoding=”REQUEST-ENCODING”?><cpse:request xmlns:cpse=”www.clusterpoint.com”><cpse:storage>storage name</cpse:storage><cpse:command>command name</cpse:command><cpse:timestamp>message date and time</cpse:timestamp><cpse:requestid>message number</cpse:requestid><cpse:application>creator of message</cpse:application><cpse:user>user name</cpse:user><cpse:password>user password</cpse:password><cpse:reply_charset>reply encoding</cpse:reply_charset><cpse:content>
</cpse:content></cpse:request>
Search
<query> search query </query><docs> number of documents </docs><offset> intend from the beginning </offset><case_sensitive> boolean type parameter</case_sensitive><relevance> boolean type parameter</relevance><group_size> maximum from one group </group_size><rate_from> FROM value </rate_from><rete_to> TO value </rate_to>
Insert
<document><id>document id</id><title>document title</title><rate>document rate</rate><info>meta data</info><site>document id</site><text>textual information</text><hidden>information that is not shown</hidden></document>
Lookup
<document><id>document id</id></document>
Platform
Runs on *nix (tested on Linux and FreeBSD)
Written in C/C++
Optimized for multi-core processorsSource code is IP of Clusterpoint
written from the scratch
PORTS
Data tcp: 5550, 80
Unix domains sockets
Cluster discovery UDP: 234.25.25.25:5550 21.04.23
ParametersDisk space
1.,5-2 times more than disk space Data of 100 GB = 150-200 GB The amount doesn’t include space for log files, as its
possible rotate and backup files, While file load and indexing size can increase 3-4
times, then return to normal size
RAM more RAM - more cached data –better
performance usually recomended >4 GB
21.04.23
Use
ComplementarySolving performance issues and bottlenecks of existing database systems
StandaloneApplication is implemented using Clusterpoint DBMS
21.04.23
APP server
SQL
USERS
Clusterpoint XML
DBMS
XML
APP server
USERS
THANK YOU
21.04.23