DefaultHashedDataSourceCluster (Apache Excalibur API)

Overview

Package

Class

Use

Tree

Deprecated

Index

Help

PREV CLASS NEXT CLASS

FRAMES NO FRAMES

SUMMARY: NESTED | FIELD | CONSTR | METHOD

DETAIL: FIELD | CONSTR | METHOD

org.apache.avalon.excalibur.datasource.cluster
Class DefaultHashedDataSourceCluster

java.lang.Object
  org.apache.avalon.framework.logger.AbstractLogEnabled
      org.apache.avalon.excalibur.datasource.cluster.AbstractDataSourceCluster
          org.apache.avalon.excalibur.datasource.cluster.DefaultHashedDataSourceCluster

All Implemented Interfaces:: Component, Configurable, DataSourceComponent, Disposable, HashedDataSourceCluster, Initializable, LogEnabled, Serviceable, ThreadSafe

public class DefaultHashedDataSourceCluster
extends AbstractDataSourceCluster
implements HashedDataSourceCluster

The DefaultHashedDataSourceCluster requires that the user specify an object or a hashCode which will be used consistantly select a member DataSource form a cluster for each connection request. Calls to getConnection() will throw an exception. Components which make use of this class must call either the getConnectionForHashObject( Object hashObject) or the getConnectionForHashCode( int hashCode ) methods instead.

This form of Clustering is useful in cases where data can be reliably accessed in a repeatable manner. For example a web site's visitor information could be accessed by using a String containing the visitor's username as a hashObject. This would allow visitor information to be spread across several database servers.

The Configuration for a 2 database cluster is like this:

   <datasources>
     <hashed-cluster name="mydb-cluster" size="2">
       <dbpool index="0">mydb-0</dbpool>
       <dbpool index="1">mydb-1</dbpool>
     </hashed-cluster>
   </datasources>
   <cluster-datasources>
     <jdbc name="mydb-0">
       <pool-controller min="1" max="10"/>
       <auto-commit>true</auto-commit>
       <driver>com.database.jdbc.JdbcDriver</driver>
       <dburl>jdbc:driver://host0/mydb</dburl>
       <user>username</user>
       <password>password</password>
     </jdbc>
     <jdbc name="mydb-1">
       <pool-controller min="1" max="10"/>
       <auto-commit>true</auto-commit>
       <driver>com.database.jdbc.JdbcDriver</driver>
       <dburl>jdbc:driver://host1/mydb</dburl>
       <user>username</user>
       <password>password</password>
     </jdbc>
   </cluster-datasources>

With the following roles declaration:

   <role name="org.apache.avalon.excalibur.datasource.DataSourceComponentSelector"
     shorthand="datasources"
     default-class="org.apache.avalon.excalibur.component.ExcaliburComponentSelector">
     <hint shorthand="jdbc" class="org.apache.avalon.excalibur.datasource.JdbcDataSource"/>
     <hint shorthand="j2ee" class="org.apache.avalon.excalibur.datasource.J2eeDataSource"/>
     <hint shorthand="hashed-cluster"
       class="org.apache.avalon.excalibur.datasource.cluster.DefaultHashedDataSourceCluster"/>
   </role>
   <role name="org.apache.avalon.excalibur.datasource.DataSourceComponentClusterSelector"
       shorthand="cluster-datasources"
       default-class="org.apache.avalon.excalibur.component.ExcaliburComponentSelector">
     <hint shorthand="jdbc" class="org.apache.avalon.excalibur.datasource.JdbcDataSource"/>
     <hint shorthand="j2ee" class="org.apache.avalon.excalibur.datasource.J2eeDataSource"/>
   </role>

A hashed-cluster definition enforces that the configuration specify a size. This size must equal the number of datasources referenced as being members of the cluster. Any datasource can be a member of the cluster.

The hashed-cluster can be obtained in the same manner as a non-clustered datasource. The only difference is in how it is used. The HashedDataSourceCluster requires that the caller provide an object or a hashCode to use when requesting a connection.

The following code demonstrates a change that can be made to database enabled components so that they will be able to work with both HashedDataSourceCluster DataSources and regular DataSources.

old:

   Connection connection = m_dataSource.getConnection();

new:

   Connection connection;
   if ( m_dataSource instanceof HashedDataSourceCluster )
   {
     connection = ((HashedDataSourceCluster)m_dataSource).getConnectionForHashObject( hashObject );
   }
   else
   {
     connection = m_dataSource.getConnection();
   }

Since:: 4.1
Version:: CVS $Revision: 1.4 $ $Date: 2004/02/28 11:47:20 $
Author:: Avalon Development Team

Field Summary

Fields inherited from class org.apache.avalon.excalibur.datasource.cluster.AbstractDataSourceCluster

m_size

Fields inherited from interface org.apache.avalon.excalibur.datasource.cluster.HashedDataSourceCluster

ROLE

Constructor Summary
`DefaultHashedDataSourceCluster()`

Method Summary
`java.sql.Connection`	`getConnection()` Not supported in this component.
`java.sql.Connection`	`getConnectionForHashCode(int hashCode)` Gets a Connection to a database given a hash code.
`java.sql.Connection`	`getConnectionForHashObject(java.lang.Object hashObject)` Gets a Connection to a database given a hash object.
`int`	`getIndexForHashCode(int hashCode)` Gets the index which will be resolved for a given hashCode.
`int`	`getIndexForHashObject(java.lang.Object hashObject)` Gets the index which will be resolved for a given hashCode.

Methods inherited from class org.apache.avalon.excalibur.datasource.cluster.AbstractDataSourceCluster

configure, dispose, getClusterSize, getConnectionForIndex, initialize, service

Methods inherited from class org.apache.avalon.framework.logger.AbstractLogEnabled

enableLogging, getLogger, setupLogger, setupLogger, setupLogger

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface org.apache.avalon.excalibur.datasource.cluster.HashedDataSourceCluster

getClusterSize, getConnectionForIndex

Methods inherited from interface org.apache.avalon.framework.configuration.Configurable

configure

Constructor Detail