CODEWITHSUNDEEP
wso2wso2-enterprise-integratorwso2-esb
Justin
Justin

Reputation: 977

WSO2 Worker Node is departed from Cluster Loop (V6.5.0)

There is Cluster Set up in our Production Environment( WSO2 Server 6.5.0 -One Manger and two Worker Node).

often we are facing ERROR Like INFO {org.wso2.carbon.ntask.core.impl.clustered.ClusterGroupCommunicator} - Task [ESB_TASK] member departed [Member [Our Worker Node IP]:4100], rescheduling missing tasks... in Manager Node,

which means Worker Node is departed from Cluster Loop. During this time, as temporary solution, we manually restart failed node, after some time (3 to 5 min), Departed Node is re-joined into Cluster loop.

Is there any permanent fix for this issue?

Awaiting for your response experts!!

Upvotes: 1

Views: 85

Answers (1)

Dilan Premarathna
Dilan Premarathna

Reputation: 1294

This seems to be an unusual issue, based on the given information I can suggest you to try and adjust the hazelcast properties and see if it helps. The hazelcast.properties file is available in <EI_HOME>/conf/hazelcast.properties. If this is not available you can create a configuration file and add the following to this file.

hazelcast.heartbeat.interval.seconds=1
hazelcast.master.confirmation.interval.seconds=5
hazelcast.max.no.heartbeat.seconds=45
hazelcast.max.no.master.confirmation.seconds=60

Hazelcast uses heartbeat monitoring as a fault detection mechanism. There, the nodes send heartbeats to other nodes. If a heartbeat message is not received by a given amount of time, then Hazelcast assumes the node is dead. This is configured via the above properties.

If you are unable to resolve the issue even with this configuration it is better to validate the clustering configuration.

Upvotes: 1

Related Questions