none
Disk space disappearing - VSS? RRS feed

  • Question

  • Hi,

    Disk space is being eaten up quite quickly on my CSVs and I can't see where its going.  I think it might be VSS but how do I tell?

    My environment is Server 2012 cluster with iSCSI storage on Dell EqualLogic holding the CSVs.  It is being backed up by DPM2012 SP1 RU1.

    I am getting diskspace warnings from the SAN that don't match up with whats on the disk.  On a 3TB CSV, Explorer tells me there is 1.81 TB in use, Cluster manager tells me there is 777GB free (2.22 in use) but if I do properties on the volume it says 329GB free space which matches what the SAN is reporting.

    How can I see what space is being used by VSS, and how can I recover some of this space, assuming it is DPM / VSS that is using it?

    Cheers,

    James

    Sunday, February 10, 2013 4:47 AM

All replies

  • Hi,

    So you have a single CSV ?  Are you using a VSS hardware provider ?

    Not sure if this works on CSV, it should, run: Vssadmin list shadowstorage

    That will list the MAXIMUM amount of space shadows can use (should be unbounded), the allocated space and more importantly, the used space for shadows for each volume on the host when using the Microsoft VSS software provider.


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. Regards, Mike J. [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.

    Sunday, February 10, 2013 4:24 PM
    Moderator
  • Hi Mike,

    Thanks for the quick response.

    I have 4 CSVs attached to this cluster, only one of them is showing strange growth.  I don't have any hardware VSS providers installed, Im using the Microsoft VSS provider.

    vssadmin list shadowstorage
    vssadmin 1.1 - Volume Shadow Copy Service administrative command-line tool
    (C) Copyright 2001-2012 Microsoft Corp.

    No items found that satisfy the query.

    vssadmin list shadows
    vssadmin 1.1 - Volume Shadow Copy Service administrative command-line tool
    (C) Copyright 2001-2012 Microsoft Corp.

    Contents of shadow copy set ID: {4a0c5c80-6388-40a9-98d5-7c955f589c50}
       Contained 1 shadow copies at creation time: 25/01/2013 10:03:00 PM
          Shadow Copy ID: {1440197b-9253-4879-8a88-b1660a2da8e3}
             Original Volume: (?)\\?\GLOBALROOT\Device\HarddiskVolume29\
             Shadow Copy Volume: \\?\GLOBALROOT\Device\HarddiskVolumeShadowCopy40
             Originating Machine: HPV02.CLOUD.Local
             Service Machine: HPV02.CLOUD.Local
             Provider: 'Microsoft Software Shadow Copy provider 1.0'
             Type: ClientAccessible
             Attributes: Persistent, Client-accessible, No auto release, No writers, Differential

    I've seen a few other people comment on similar issues here: http://social.technet.microsoft.com/Forums/en-US/dpmhypervbackup/thread/44cde22c-e404-485d-8e71-246525204208/ and here: http://www.starwindsoftware.com/forums/starwind-f5/dpm2012-csv-and-hardware-vss-t3009-15.html "I'm having the same issue with the snapshots piling up."

    Its an issue from growing my dynamically expanding SAN volumes unexpectedly and also I'm concerned about what will happen when the SAN volume gets full - will it pause my VMs or will my backups start failing?

    Does Microsoft have any patches or fixes for this on the way?

    Cheers,

    James

    Sunday, February 10, 2013 8:30 PM
  • Hi,

    The snapshot listed above was created  25/01/2013 10:03:00 PM and is of Type: ClientAccessible.  That would not be created for a hyper-v guest backup.  See what the other nodes show for shadow copies.   When DPM makes a shadow copy for VM backup, the snapshot type would be Type: ApplicationRollback


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. Regards, Mike J. [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.

    Monday, February 11, 2013 3:22 PM
    Moderator
  • Hi,

    We have got the same issue here.

    Hyper-V Cluster 2012 with 4 CSV. On Volume 2 which is 1.3 TB in size we have 8 virtual machines. (dynamic disk) The sum size of these machine are 660GB, but the cluster manager show only 60!! GB free space.

    We use DPM 2012 to back up VMs.

    My question is: Where is the remaining 600 GB????

    We've run the vssadmin commands but with no result!

    We have no snapshots. No volume snapshots.

    Anyone can help?


    Tuesday, March 3, 2015 3:56 PM
  • Hi Gergely,

    In my environment this issue was resolved after rebooting each of the nodes and the issue has not re-occurred.

    Tuesday, March 3, 2015 8:26 PM
  • Hi James,

    Thank you for your response.

    We rebooted all nodes but with no luck. The space is still missing:(


    Wednesday, March 25, 2015 12:47 PM
  • We typically will move the ownership between nodes and the space is recovered.
    • Proposed as answer by BSUR Saturday, April 30, 2016 3:45 PM
    Thursday, May 21, 2015 4:16 PM
  • Recently we have deployed VEEAM B&R v8 to protect our Hyper-V 2012 R2 Cluster and this issue presented itself on one of the 2 CSVs we use for production...  Doing exactly what Charlotte has suggested resolved the missing space.

    Curious if it would be worth the effort to write up a PS script to move the CSV to a different owner after the VEEAM backup completes...

    Wednesday, July 1, 2015 6:28 PM
  • Hi all,

    We also had this issue with our CSV volume showing up as near full, but when you check the file size there is hardly anything in the CSV folder.

    We are running Veeam Version 8 and Server 2012 R2, moving the CSV to another Node did reclaim the space but does anyone know of a hotfix for this issue?

    Monday, August 24, 2015 4:07 AM
  • Moving the ownerships (Move to best node) did magically solve the issue, regained a Few TB by moving csv's..
    Saturday, April 30, 2016 3:45 PM