Dual active Java Message Service (JMS) brokers ensure that there is no loss of service if a primary JMS broker is down. The files needed on both the local and remote sites to configure dual active JMS brokers are as follows:
- daemon.properties
- agent.properties
- commandline.properties
- Log onto the local daemon host and run the following command as root: ./dmcluster configactivemq -e true -s slaveDaemonHost-p 61616
-
Log on to the remote daemon host and run the following command as root:
./dmcluster configactivemq -e true -s masterDaemonHost -p 61616
Where:
Designation Description -e Enable network of brokers configuration -s Remote host/server name where the other ActiveMQ instance is running -p Port to connect to the remote ActiveMQ - Inspect TDActiveMQ logs to ensure no errors/warnings are present: /var/opt/teradata/tdactivemq/logs/activemq.log
- Enter cluster.enabled=true in daemon.properties, agent.properties, and commandline.properties on the local and remote sites to enable connection to a secondary JMS broker.
- Edit the broker.url property in daemon.properties and commandline.properties on the local and remote sites to specify a secondary JMS host. For example, broker.url=primaryJmshost, secondaryJmsHost On the local site, the primaryJmshost will be the master daemon host and the secondaryJmsHost will be the slave daemon host. On the remote site, the primaryJmshost will be the slave daemon host and the secondaryJmsHost will be the master daemon host.
- [Optional] Run the following command on the local and remote daemon hosts to restore the ActiveMQ to the standard, non-cluster configuration: ./dmcluster configactivemq -e false
- [Optional] Inspect TDActiveMQ logs to ensure that TDActiveMQ is restarted in a standard (non-cluster) configuration with no errors.
- Edit the broker.url property in agent.properties on the local and remote sites to specify a secondary JMS host. For example, broker.url=primaryJmshost, secondaryJmsHost Make sure that the order is the same for all agents, regardless of whether they are considered master or slave. For example, all agents will have broker.url=primaryJmshost, secondaryJmsHost. The primaryJmshost and secondaryJmsHost do not change based on site location. If the order is not the same, Data Mover job functionality might be impacted.
- Select Enable in portlets and enter the broker URL/port value for the secondary JMS broker in the Data Mover Setup portlet to enable clustering when connecting with the daemon. For example, dm-agent:61616, where dm-agent is the broker URL and 61616 is the port number. The clustering field values must be set individually for each daemon.