[AMQ-6953] Active MQ messages lost after failover - pacemaker

• Type: Task
• Status: Closed
• Priority: Major
• Resolution: Invalid
• Affects Version/s: 5.15.2
• Fix Version/s: NEEDS_REVIEW
• Component/s: Broker
• Assignee: Unassigned
• Reporter: Faheem Khan
• Created: 26/Apr/18 12:25
• Updated: 26/Apr/18 13:17
• Resolved: 26/Apr/18 13:17

Description

I have configured high available ActiveMQ cluster setup where ActiveMQ is deployed on WebLogic server.
I have configured corosync and pacemaker for failover through VIP.
The problem I am facing is when I send 10 messages to active node and shut it down, on the secondary node I didn't see any messages recovered, but when I shutdown secondary node messages recovered on primary Node, I have shared NFS of entire mount point and shared kahadb logs, where the application is installed on both nodes.

Seems like messages are stick with each connection, i check logs and found that each server is creating its own ClientID

e.g  Apache ActiveMQ 5.12.1 (localhost, ID:NDL-WLS-301.domain.com-48501-1517840420812
-0:1)

Host-Primery: NDL-WLS-301

Host-Secondery: NDL-WLS-302

please help