VM with wrong Service Offering avec host went into maintenance mode ?

[Fraysse Guillaume <gf...@gmail.com>]

Hi to the list,

A customer notified us that one of its VM is wrongly displayed in CS.
This VM shows up using a service offering of 1 vCPU and 1 GB of RAM
when it is a 4vCPU and 8GB VM in vCenter, which is confirmed in the OS
itself.

Context :
- CloudStack 2.2.13
- VMware ESXi 5.0
- VM is a "classic" Debian 6, same template is used for hundreds of VM

Looking back at CloudStack logs, I only saw that line (I anonymised
some of the data) :
2013-03-06 09:44:46,058 DEBUG [agent.transport.Request]
(HA-Worker-4:work-6557) Seq 4-1829307871: Executing: { Cmd , MgmtId:
345051320510, via: 4, Ver: v1, Flags: 100111,
[{"PrepareForMigrationCommand":{"vm":{"id":377,"name":"i-56-377-VM","bootloader":"HVM","type":"User","cpus":1,"speed":1000,"minRam":1073741824,"maxRam":1073741824,"arch":"x86_64","os":"Debian
GNU/Linux 6(64-bit)","bootArgs":"","rebootOnCrash":false,"enableHA":false,"limitCpuUse":false,"vncPassword":"4f2dd6c2f15ef53f","params":{"nic_adapter":"E1000"},"disks":[{"id":486,"name":"ROOT-377","mountPoint":"/primarystorage","path":"ROOT-377","size":9663676416,"type":"ROOT","storagePoolType":"NetworkFilesystem","storagePoolUuid":"d9ddfc82-c3a8-311f-b8c6-57b765af0630","deviceId":0},{"id":487,"name":"DATA-377","mountPoint":"/primarystorage","path":"90d3411ce87049b68863a53f88ce976b","size":53687091200,"type":"DATADISK","storagePoolType":"NetworkFilesystem","storagePoolUuid":"d9ddfc82-c3a8-311f-b8c6-57b765af0630","deviceId":1}],"nics":[{"deviceId":0,"networkRateMbps":200,"defaultNic":true,"ip":"10.200.64.6","netmask":"255.255.255.192","gateway":"10.200.64.1","mac":"06:4b:94:00:16:1b","dns1":"10.200.193.1","dns2":"10.94.53.3","broadcastType":"Vlan","type":"Guest","broadcastUri":"vlan://3052","isolationUri":"vlan://3052","isSecurityGroupEnabled":false,"tags":[]},{"deviceId":1,"networkRateMbps":200,"defaultNic":false,"ip":"10.200.132.208","netmask":"255.255.255.128","gateway":"10.200.132.193","mac":"06:2c:8e:00:16:36","dns1":"10.200.193.1","dns2":"10.94.53.3","broadcastType":"Vlan","type":"Guest","broadcastUri":"vlan://3053","isolationUri":"vlan://3053","isSecurityGroupEnabled":false,"tags":[]},{"deviceId":2,"networkRateMbps":200,"defaultNic":false,"ip":"10.200.0.6","netmask":"255.255.255.192","gateway":"10.200.0.1","mac":"06:76:3a:00:16:00","dns1":"10.200.193.1","dns2":"10.94.53.3","broadcastType":"Vlan","type":"Guest","broadcastUri":"vlan://3051","isolationUri":"vlan://3051","isSecurityGroupEnabled":false,"tags":[]}]},"wait":0}}]
}

There was no change of Service Offering of this VM, and even if there
was one we weren't aware of, it is consistant as the VM would have
been rebooted and would have show up consistantly in VMware and
CloudStack.

The above log seems to date to a maintenance operation where we put
one of the host maintenance mode.

Has anyone experienced this before ? I haven't found a bug like this
in Jira or in the list.

Best regards,
Guillaume