During host preparation in VMware NSX for vSphere 6.x, you experience these symptoms:
The vSphere ESX Agent Manager (EAM) threads get stuck or EAM service may not start.
In the C:\ProgramData\VMware\VMware VirtualCenter\Logs\eam.log file on the vCenter Server, you see entries similar to:
2015-04-23T06:31:38.998Z | DEBUG | agent-4 | AgentImpl.java | 2251 | Agent AgentImpl(ID:'Agent:e83ee5ce-####-####-####-###########20:3c36c535-####-####-####-##########11') (VcHostSystem(ID: host-359)) scheduled a VIB task and is waiting for it to complete. 2015-04-23T06:31:40.379Z | DEBUG | compute-res-0 | ExecutorImpl.java | 57 | JOB SUBMITTED: [#1420322380] ClusterVibConditionalJob(ManagedObjectReference: type = ClusterComputeResource, value = domain-c343, serverGuid = 885ab205-####-####-####-##########79), trace: .:scheduleNewVibTask:. 2015-04-23T06:31:40.379Z | DEBUG | compute-res-0 | AuditedJob.java | 70 | JOB STARTED: [#1420322380] ClusterVibConditionalJob(ManagedObjectReference: type = ClusterComputeResource, value = domain-c343, serverGuid = 885ab205-####-####-####-##########79), trace: .:scheduleNewVibTask:.
INFO | jvm 1 | 2015/04/23 06:32:20 | java.lang.OutOfMemoryError: Java heap space INFO | jvm 1 | 2015/04/23 06:32:20 | Dumping heap to /var/log/vmware/eam/java_pid24722.hprof ... INFO | jvm 1 | 2015/04/23 06:32:22 | Heap dump file created [138746180 bytes in 2.438 secs] INFO | jvm 1 | 2015/04/23 06:32:22 | Exception in thread "host-359-0" java.lang.OutOfMemoryError: Java heap space INFO | jvm 1 | 2015/04/23 06:32:22 | at sun.nio.cs.StreamEncoder.write(Unknown Source) INFO | jvm 1 | 2015/04/23 06:32:22 | at java.io.OutputStreamWriter.write(Unknown Source) INFO | jvm 1 | 2015/04/23 06:32:22 | at java.io.Writer.write(Unknown Source)
INFO | jvm 1 | 2015/04/23 06:32:23 | at com.vmware.eam.vc.VcListener$4.call(VcListener.java:338) INFO | jvm 1 | 2015/04/23 06:32:23 | at com.vmware.eam.vc.VcListener$4.call(VcListener.java:335) INFO | jvm 1 | 2015/04/23 06:32:23 | at com.vmware.eam.lock.RaceProtection.exec(RaceProtection.java:68) INFO | jvm 1 | 2015/04/23 06:32:23 | at com.vmware.eam.lock.RaceProtection.exec(RaceProtection.java:51) INFO | jvm 1 | 2015/04/23 06:32:23 | at com.vmware.eam.vc.VcListener.unregisterObserver(VcListener.java:335)
2015-04-23T06:35:43.033Z | DEBUG | agent-3 | AgentImpl.java | 2251 | Agent AgentImpl(ID:'Agent:86cab9ca-####-####-####-##########a5:3c36c535-####-####-####-##########11') (VcHostSystem(ID: host-363)) scheduled a VIB task and is waiting for it to complete.
Note: The preceding log excerpts are only examples. Date, time, and environmental variables may vary depending on your environment.
Environment
VMware vCenter Server 6.0.x VMware NSX for vSphere 6.0.x VMware vCenter Server Appliance 6.0.x VMware NSX for vSphere 6.1.x VMware NSX for vSphere 6.2.x
Cause
This issue occurs due to the bug in EAM where it tries to poll the same updates again from VM exhausting the memory of the EAM.
Resolution
This issue is resolved in:
vCenter Server 6.0 Update 3
vCenter Server 6.5 Update 1
To work around the issue, increase the EAM heap size to the desired level using cloudvm-ram-size utility.