VMware Cloud Community
Travis_Liu
Contributor
Contributor
‎08-13-2013 02:38 AM

Cannot power on VM.

Hi All,

Recently I always encounter the issue that one or several VMs cannot be powered on after deployment. In the VMWare.log it will contains the following error message:

VMWare.log

Aug 12 13:49:41.175: vmx| DISKLIB-DSCPTR: Opened [0]: "021966-Win2K8R2SP1_80G_Clean-0-0-scsi-flat.vmdk" (0xe)

Aug 12 13:49:41.175: vmx| DISKLIB-LINK  : Opened '/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/21966/021966-Win2K8R2SP1_80G_Clean-0-0-scsi.vmdk' (0xe): vmfs, 167772160 sectors / 80 GB.

Aug 12 13:49:41.186: vmx| DISKLIB-CHAINESX : ChainESXOpenSubChain: numLinks = 13, numSubChains = 1

Aug 12 13:49:41.419: vmx| DISKLIB-CHAINESX : ChainESXOpenSubChainNode: can't create deltadisk node 27a85124-030515-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk failed with error Not a directory

Aug 12 13:49:41.419: vmx| DISKLIB-CHAIN : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/30515/030515-Win2K8R2SP1_80G_Clean-0-0-scsi.vmdk" : failed to open (Not a directory).

Aug 12 13:49:41.420: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/30515/030515-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.421: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/30207/030207-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.422: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/30121/030121-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.423: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/24403/024403-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.425: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/23713/023713-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.426: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/22102/022102-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.427: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/22075/022075-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.428: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/22050/022050-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.428: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/21998/021998-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.428: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/21979/021979-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.428: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/21978/021978-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.428: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/21973/021973-Win2K8R2SP1_80G_Clean-0-0-scsi-delta.vmdk" : closed.

Aug 12 13:49:41.428: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/21966/021966-Win2K8R2SP1_80G_Clean-0-0-scsi-flat.vmdk" : closed.

Aug 12 13:49:41.429: vmx| DISKLIB-LIB   : Failed to open '/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/30515/030515-Win2K8R2SP1_80G_Clean-0-0-scsi.vmdk' with flags 0xa (Not a directory).

Aug 12 13:49:41.429: vmx| DISK: Cannot open disk "/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/30515/030515-Win2K8R2SP1_80G_Clean-0-0-scsi.vmdk": Not a directory (1310729).

Aug 12 13:49:41.429: vmx| Msg_Post: Error

Aug 12 13:49:41.429: vmx| [msg.disk.noBackEnd] Cannot open the disk '/vmfs/volumes/4f2b7294-6a77fd3f-f49a-0024e82e2501/saaslm/30515/030515-Win2K8R2SP1_80G_Clean-0-0-scsi.vmdk' or one of the snapshot disks it depends on.

Aug 12 13:49:41.429: vmx| [msg.disk.configureDiskError] Reason: Not a directory.----------------------------------------

Aug 12 13:49:41.467: vmx| Module DiskEarly power on failed.

Aug 12 13:49:41.467: vmx| VMX_PowerOn: ModuleTable_PowerOn = 0

VCenter 4.0.0, 208111.

I searched here and did not found any similar issue with the error "Not a directory", dose anyone have some ideas about this? Thanks in advance.

Travis

0 Kudos
5 Replies
a_p_
Leadership
Leadership
‎08-13-2013 02:49 AM

Welcome to the Community,

How did you deploy the VMs. The entries look kind of weird, i.e. multiple delta.vmdk files instead of a flat.vmdk file!? Please compress/zip the VM's vmware.log file as well as the small .vmdk descriptor text file(s) and attach the .zip file to a reply post. To attach files you need to use the advanced editor (link on the upper right in the edit box).

André

Travis_Liu
Contributor
Contributor
‎08-13-2013 03:06 AM

Hi André,

Thanks a lot for your response. I attached the log here.

Travis

vmware.log
031552-Win2K8R2SP1_80G_Clean-0-0-scsi.vmdk
0 Kudos
a_p_
Leadership
Leadership
‎08-13-2013 03:35 AM

According to the log files you are deploying a VM with a delta (snapshot) disks, which has another 6 links (snapshots in the chain)!? To me this doesn't look like it is really intended to be this way. Can you please explain the purpose for this i.e. why it is done this way. This doesn't only decrease performance, but also is risky because you will loose the VM if only a single chain-link gets lost.

André

0 Kudos
Respen33
Contributor
Contributor
‎08-13-2013 04:56 AM

I agree with Andre in that it looks like you have nested snapshots on this VM. You may want to consolidate the snapshots back into the primary VMDK (through snapshot manager) then try again to see if the result changes for you. Each snapshot will create a delta VMDK i believe which the primary image will reference, kind of like a pointer to data instead of transferring data.

Often in error, never in doubt.
Travis_Liu
Contributor
Contributor
‎08-13-2013 11:44 PM

Thanks all for your reply. Actually, my lab environment is a lab manager environment, the multiple delta disks may caused by lab manager create configuration from template, and clone to library then clone to configuration from the library and so on. So it is not easy for us to consolidate the delta back to the primary one. One phenomenon I observed is, this issue usually occurs when there are lots of VMs deployed on one single server, and when the issue occurs, try to un-deploy several VMs could often resolve the problem, I am not sure if there are any limitation inside the system, or there is some kind of system defect occurs when specific resource is exhausted. We sometimes really feel frustrated with it.

0 Kudos