hello guys,
i have one problem with one vm, which is having snapshots...actully what i did is i have cold migrated this vm and now when i try to start the vm it gives me error stating internal error..
when i checked the datastore for vm files it show me .delta.vmdk files and my virtul disk is pointing towards my diskname.0000001.vmdk which should actually point to diskname.vmdk...
i m using ESX version 3.5
i have tried to remove snapshot through console but this gives me error vm not found, i think because the system is offline....
can anyone help me with exact solution....
Hello,
Sounds like the .vmsd file is out of sorts. Look at the vmware.log file for the VM for a better error message. That will be invaluable.
Best regards,
Edward L. Haletky
VMware Communities User Moderator
====
Author of the book 'VMWare ESX Server in the Enterprise: Planning and Securing Virtualization Servers', Copyright 2008 Pearson Education.
CIO Virtualization Blog: http://www.cio.com/blog/index/topic/168354
As well as the Virtualization Wiki at http://www.astroarch.com/wiki/index.php/Virtualization
Hello,
This issue is not difficult to correct.
When you take a snapshot all the existing vmdk's become readonly references and a delta is created for all the writes.
The vmx file held the info for the current active vmdk before a snapshot as shown here.
scsi0:0.present = "true"
scsi0:0.fileName = "VD1.vmdk"
scsi0:0.deviceType = "scsi-hardDisk"
And then after a snapshot
scsi0:0.present = "true"
scsi0:0.fileName = "VD1-000001.vmdk"
scsi0:0.deviceType = "scsi-hardDisk"
You will need to point it back to the original vmdk as it was before a snapshot was initiated and the info can be observed in the vmware logs.
Also since you moved the VM to a new store the vswp file will be invalid.
This is not an issue all you need to do is clear the entry for it in the vmx file.
e.g.
sched.swap.derivedName = "/vmfs/volumes/486b0a84-b380612a-2919-000d60d46e2e/VD1/VD1-92f3a8ab.vswp"
must be edited to
sched.swap.derivedName = ""
and it will gen a new one on power up.
If multiple snaps were present then more work will need to be done and addition components come into play.
I would hope that you would not leave snapshots on VM's for any time longer than required to back them up.
Hello Mike,
thanks for the valuable information, sure i will not keep the snapshots for longer...
there are multiple snapshots, could u pls let me know how to clear this and get rid out of this.....
Thanks
Amit
Your welcome, a great way to say thanks is to award points.
When there are multiple snapshots we need all the vmdk descriptor files and delta's except the last one to be present in the VM directory.
Lets say there are three snapshots.
We would point the vmx to VMname-000002.vmdk in this case.
The original would be.
scsi0:0.fileName = "VD1.vmdk"
The first snap
scsi0:0.fileName = "VD1-000001.vmdk"
Second wee need to point it here
scsi0:0.fileName = "VD1-000002.vmdk"
Third snap we will disgard
scsi0:0.fileName = "VD1-000003.vmdk"
This activity is observable in the vmware logs
One thing you must be aware of is if a snap was deleted before we got to this point sometime in the past the sequence changes.
If the second snap got deleted then the last pointer would be
scsi0:0.fileName = "VD1-000003.vmdk"
But we would need to point to
scsi0:0.fileName = "VD1-000001.vmdk"
Instead of 2 which is missing.
You should remove these lines.
scsi0:0.present = "true"
scsi0:0.fileName = "sbm01.vmdk"
scsi0:0.deviceType = "scsi-hardDisk"
scsi0:1.present = "true"
scsi0:1.fileName = "sbm01_1.vmdk"
scsi0:1.deviceType = "scsi-hardDisk"
What vmdk files do you have in VM directory?
Can you post the last couple of vmware.log's when it was snapshotted?
Mike,
Actually this is
scsi0:0.present = "true"
scsi0:0.fileName = "sbm01-000001.vmdk"
scsi0:0.deviceType = "scsi-hardDisk"
scsi0:1.present = "true"
scsi0:1.fileName = "sbm01_1-000001.vmdk"
scsi0:1.deviceType = "scsi-hardDisk"
and the vm is not booting...
according to our earlier disscussion i have replaced this, so as it should point the original vmdk . i.e: sbm01.vmdk
i have attached the required log file.
Thanks,
Amit
Since May 21/08 you have been running on snapshot file instance 000001.
So there is only one snapshot present.
Was a backup made while the VM was powered on and do you want to go to the May 21/08 point in time?
I am not sure about the backup and i don't wnat to go back...what i want at present is to get my machin up and running
Thanks,
Amit
I will need more info to tell you how to proceed.
1. Can you post a directory listing of the files in the VM folder.
use
ls -l > dirlist.txt
2. Can you upload the vmsd files the vmdk descriptors.
I am sorry mike there is nothing such in the dir list
actually one of our team member has build the new vm and attached the vmdks and started the vm..
Thanks for your prompt response eachtime....
Is there anything you can suggest to recover from this if in case i face the same again?
Thanks,
Amit