Snapshot disks remain after crash

[mattmiller1505]

I have a Windows server that has crashed a couple of times during a backup with the Data recovery tool.  During both crashes, the server would not restart unless I deleted the VDR snapshot while the server was powered off. Now within my datastore for this server I see multiple instances of the disks, and I assume these are remnants of the snaphot.

sqldb_1.vmdk     Size: 200GB, Provisioned: 0

sqldb_1-00001.vmdk     Size: 58 GB, Provisioned 200GB

sqldb_1-00002.vmdk     Size 313 MB, Provisioned 200GB

i don't trust the reported sizes on these disks, There is about 130GB of actual data on that disk.  How can I tell which disk is active? Or is it somehow using all 3? Is it safe to just delete the non-used disks?

Thanks

[a_p_]

Welcome to the Community,

the best way would be to take a look at the latest vmware.log as well as the VM's .vmx file. If the .vmx file shows the base .vmdk file and the snapshots are not mentioned in the vmware.log, it should be safe to delete these files. It might be a good idea though to create s sub-directory and move the snapshot files to this folder and delete them later. You may also want to move/delete the .vmsd as well as the .vmsn files to clean up the folder.

If you like, post a list of files in the VM's folder and attach the vmware.log and the .vmx file to your next post. This would help to give you a recommendation.

André

[mattmiller1505]

Thanks for the insight.  It looks to me the active harddrives are sqldb3-000001.vmdk (disk 1) and sqldb3_1-000002.vmdk (disk 2).  If that is correct, I will move all other vmdks to a subfold for a few days to make sure they are not needed and then delete them, as you suggest.

Can you help me understand the Size information that is displayed on the datastore picture? All disks are thick provisioned, and both drives contain more data than what is listed in the Size column.  This obviously isn't a size on disk or disc consumption.

[a_p_]

Don't delete/move any files manually! Currently all .vmdk files are in use!

Snapshots in VMware products are used like a chain. Currently - according to the log file - both virtual disks run on snapshots, where sqldb3.vmdk has 1 snapshot and sqldb3_1.vmdk has two of them. Unless you need these snapshots, open the "Snapshot Manager" by right clicking the VM, create a new snapshot and then select "Delete All" (after working hours).

André

[a_p_ · Accepted solution]

This is how the snapshot chain for the two virtual disks currently looks like.

• scsi0:0.fileName = sqldb3-000001.vmdk -> sqldb3.vmdk
• scsi0:1.fileName = sqldb3_1-000002.vmdk -> sqldb3_1-000001.vmdk --> sqldb3_1.vmdk

Please take a look at http://kb.vmware.com/kb/1015180 (Understanding virtual machine snapshots in VMware ESX)

André

[a_p_]

Here you can see the parts from the vmware.log file which show the  snapshot chain:

Dec 28 01:22:27.921: vmx| DISK: OPEN scsi0:0 '/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3-000001.vmdk' persistent R[]
Dec 28 01:22:27.962: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3-000001-delta.vmdk" : open successful (10) size = 486625280, hd = 228525661. Type 8
Dec 28 01:22:27.962: vmx| DISKLIB-DSCPTR: Opened [0]: "sqldb3-000001-delta.vmdk" (0xa)
Dec 28 01:22:27.963: vmx| DISKLIB-LINK  : Opened '/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3-000001.vmdk' (0xa): vmfsSparse, 83923560 sectors / 40.0 GB.
Dec 28 01:22:27.967: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3-flat.vmdk" : open successful (14) size = 42968862720, hd = 236783198. Type 3
Dec 28 01:22:27.967: vmx| DISKLIB-DSCPTR: Opened [0]: "sqldb3-flat.vmdk" (0xe)
Dec 28 01:22:27.967: vmx| DISKLIB-LINK  : Opened '/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3.vmdk' (0xe): vmfs, 83923560 sectors / 40.0 GB.
Dec 28 01:22:27.967: vmx| DISKLIB-CHAINESX : ChainESXOpenSubChain: numLinks = 2, numSubChains = 1
...
Dec 28 01:22:27.987: vmx| DISKLIB-LIB   : Opened "/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3-000001.vmdk" (flags 0xa, type vmfs).


Dec 28 01:22:27.994: vmx| DISK: OPEN scsi0:1 '/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1-000002.vmdk' persistent R[]
Dec 28 01:22:28.008: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1-000002-delta.vmdk" : open successful (10) size = 413696, hd = 253167202. Type 8
Dec 28 01:22:28.008: vmx| DISKLIB-DSCPTR: Opened [0]: "sqldb3_1-000002-delta.vmdk" (0xa)
Dec 28 01:22:28.008: vmx| DISKLIB-LINK  : Opened '/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1-000002.vmdk' (0xa): vmfsSparse, 419430400 sectors / 200 GB.
Dec 28 01:22:28.020: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1-000001-delta.vmdk" : open successful (14) size = 58703892480, hd = 202638947. Type 8
Dec 28 01:22:28.020: vmx| DISKLIB-DSCPTR: Opened [0]: "sqldb3_1-000001-delta.vmdk" (0xe)
Dec 28 01:22:28.020: vmx| DISKLIB-LINK  : Opened '/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1-000001.vmdk' (0xe): vmfsSparse, 419430400 sectors / 200 GB.
Dec 28 01:22:28.033: vmx| DISKLIB-VMFS  : "/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1-flat.vmdk" : open successful (14) size = 214748364800, hd = 238814820. Type 3
Dec 28 01:22:28.033: vmx| DISKLIB-DSCPTR: Opened [0]: "sqldb3_1-flat.vmdk" (0xe)
Dec 28 01:22:28.033: vmx| DISKLIB-LINK  : Opened '/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1.vmdk' (0xe): vmfs, 419430400 sectors / 200 GB.
Dec 28 01:22:28.033: vmx| DISKLIB-CHAINESX : ChainESXOpenSubChain: numLinks = 3, numSubChains = 1
...
Dec 28 01:22:28.069: vmx| DISKLIB-LIB   : Opened "/vmfs/volumes/4ef8ed2d-32b06caa-9dfe-68b59977f7c2/sqldb3/sqldb3_1-000002.vmdk" (flags 0xa, type vmfs).

André

[mattmiller1505]

Thanks, that worked.  Thanks for all the valuable info on snapshots too.