{scrollbar} |
JMS clustering in Geronimo is handled by ActiveMQ component directly. |
activemq.xml
file under /var/activemq/conf
directory, you can configure brokers to be clustered and a JMS request can fail over to another broker if the JMS broker goes down using Master Slave functionality.
There are different kinds of Master/Slave configurations available according to ActiveMQ documentation:
Refer to the configuration below for each scenario in Geronimo
With this scenario, you must specify the master and slave node explicitly and manually restart a failed master
On the master node, you just need to specify the current node is a master by using brokerName attribute as followed.
... <broker xmlns="http://activemq.apache.org/schema/core" brokerName="master" useJmx="false" deleteAllMessagesOnStartup="true" tmpDataDirectory="${activemq.data}/tmp_storage" useShutdownHook="false" start="false"> ... |
Because each master only has one slave in Pure Master/Slave scenario, the slave node must know URI of master node and also be tagged as a slave node using brokerName attribute.
... <broker xmlns="http://activemq.apache.org/schema/core" brokerName="slave" deleteAllMessagesOnStartup="true" useJmx="false" masterConnectorURI="tcp://masterHostname:${${ActiveMQPort}}" tmpDataDirectory="${activemq.data}/tmp_storage" useShutdownHook="false" start="false"> ... |
JMS clients use failover:// protocol to locate brokers in a cluster such as
failover://(tcp://masterhost:61616,tcp://slavehost:61616)?randomize=false |
In this scenario, you must use a shared file system to provide high availability of brokers and automatic discovery of master/slave nodes. The shared folder must allow different slaves to have write permission.
On each node, configure a shared directory as the place where brokers are using persistenceAdapter
element as followed:
... <persistenceAdapter> <amqPersistenceAdapter directory="/sharedFileSystem/sharedBrokerData"/> </persistenceAdapter> ... |
Note that:
//ipAddress/sharedFolder
in the configuration.JMS clients use failover:// protocol to locate brokers in a cluster such as
failover://(tcp://broker1:61616,tcp://broker2:61616,tcp://broker3:61616)?randomize=false |
In this scenario, you must use a shared database as persistence enginee and automatic recovery.
On each node, configure a shared database pool using jdbcPersistenceAdapter
element as followed, we use embeded Derby database server as an example:
... <persistenceAdapter> <amqPersistenceAdapter dataSource="Shared-DS"/> </persistenceAdapter> ... <bean id="Shared-ds" class="org.apache.derby.jdbc.EmbeddedDataSource"> <property name="databaseName" value="Shared_db"/> <property name="createDatabase" value="create"/> </bean> |
JMS clients use failover:// protocol to locate brokers in a cluster such as
failover://(tcp://broker1:61616,tcp://broker2:61616,tcp://broker3:61616)?randomize=false |