[CLOUDSTACK-429] Agent rebalancing is broken

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: pre-4.0.0
• Component/s: Hypervisor Controller, Management Server
• Security Level: Public (Anyone can view this level - this is the default.)
• Environment: KVM, Cloudstack 3.0.2
• Assignee: Unassigned
• Reporter: Roland Kool
• Created: 31/Oct/12 12:16
• Updated: 17/Mar/17 16:26
• Resolved: 17/Mar/17 16:26

Description

After setting agent.lb.enabled to true and restarting my management servers, the agent rebalancing gets stuck after trying to rebalance the first 16 machines.
The error from the management log is:

2012-07-05 15:09:43,643 DEBUG [agent.manager.ClusteredAgentManagerImpl] (AgentTaskPool-6:null) Rebalancing host id=9
2012-07-05 15:09:43,646 WARN [agent.manager.ClusteredAgentManagerImpl] (AgentTaskPool-6:null) Unable to rebalance host id=9
java.lang.ClassCastException: com.cloud.agent.manager.ClusteredAgentAttache cannot be cast to com.cloud.agent.manager.ClusteredDirectAgentAttache
at com.cloud.agent.manager.ClusteredAgentManagerImpl.startRebalance(ClusteredAgentManagerImpl.java:1014)
at com.cloud.agent.manager.ClusteredAgentManagerImpl.rebalanceHost(ClusteredAgentManagerImpl.java:905)
at com.cloud.agent.manager.ClusteredAgentManagerImpl$RebalanceTask.run(ClusteredAgentManagerImpl.java:1070)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1110)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:603)
at java.lang.Thread.run(Thread.java:679)