Fuse ESB - Using Persistent Messages - Using JDBC with the High Performance Journal

Using JDBC with the High Performance Journal

Overview

Using the JDBC persistence adapter with Fuse Message Broker's high performance journal boosts the performance of the persistence adapter in two ways:

In applications where message consumers keep up with the message producers, the journal makes it possible to lower the number of messages that need to be committed to the data base. For example a message producer could publish 10,000 messages between journal checkpoints. If the message consumer pops 9,900 messages off of the queue during the same interval, only 100 messages will be committed to the database through the JDBC adapter.
In applications where the message consumers cannot keep up with the message producers, or in applications where messages must persist for long periods, the journal boosts performance by committing messages in large batches. This means that the JDBC driver can optimize the writes to the external database.

In addition to the performance gains, the high performance journal also makes it possible to ensure the consistency of JMS transactions in the case of a system failure.

Example

Example 3.3 shows a configuration fragment that configures the journaled JDBC adapter to use a MySQL database.

Example 3.3. Configuring Fuse Message Broker to use the Journaled JDBC Persistence Adapter

<beans ... >
  <broker ...>
    ...
  <persistenceFactory>
    <journaledJDBC journalLogFiles="5" dataSource="#mysql-ds" />
    </persistenceFactory>
    ...
  <broker>
  ...
<bean id="mysql-ds"
      class="org.apache.commons.dbcp.BasicDataSource"
      destroy-method="close">
    <property name="driverClassName" value="com.mysql.jdbc.Driver"/>
    <property name="url" value="jdbc:mysql://localhost/activemq?relaxAutoCommit=true"/>
    <property name="username" value="activemq"/>
    <property name="password" value="activemq"/>
    <property name="poolPreparedStatements" value="true"/>
  </bean>

The configuration in Example 3.3 has three noteworthy elements:

The persistenceFactory element wraps the configuration for the JDBC persistence adapter.

	Note
	In Fuse Message Broker versions prior to 5.4, the `persistenceAdapter` element was sometimes used to wrap the configuration for the JDBC persistence adapter. This is no longer possible in 5.4, because the broker performs strict schema checking on the configuration file.

The journaledJDBC element specifies that the broker will use the JDBC persistence adapter with the high performance journal. The element's attributes configure the following properties:

The journal will span five log files.
The configuration for the JDBC driver is specified in a bean element with the ID, mysql-ds.

The bean element specified the configuration for the MySQL JDBC driver.

Configuration

Table 3.1 describes the attributes used to configure the journaled JDBC persistence adapter.

Table 3.1. Attributes for Configuring the Journaled JDBC Persistence Adapter

Attribute	Default Value	Description
`adapter`		Specifies the strategy to use when accessing a non-supported database. For more information see Using generic JDBC providers.
`createTablesOnStartup`	`true`	Specifies whether or not new database tables are created when the broker starts. If the database tables already exist, the existing tables are reused.
`dataDirectory`	`activemq-data`	Specifies the directory into which the default Derby database writes its files.
`dataSource`	`#derby`	Specifies the id of the Spring bean storing the JDBC driver's configuration. For more information see Configuring your JDBC driver.
`journalArchiveDirectory`		Specifies the directory used to store archived journal log files.
`journalLogFiles`	`2`	Specifies the number of log files to use for storing the journal.
`journalLogFileSize`	`20MB`	Specifies the size for a journal's log file.
`journalThreadPriority`	`10`	Specifies the thread priority of the thread used for journaling.
`useDatabaseLock`	`true`	Specifies whether or not an exclusive database lock should be used to enable JDBC Master/Slave.
`useJournal`	`true`	Specifies whether or not to use the journal.

Comments powered by Disqus

Contents
Search