none
server 2012 Hyper-V CSV volume reporting incorrect space RRS feed

  • Question

  • I have a 3-node server 2012 Hyper-V cluster with 10 CSV's.

    Our main CSV (3.5TB in size holding the majority of our VM's) seems to be increasing in used space even though there have been no VM's added or any changes to it.

    A month ago this same CSV ran out of space so all the VM's ground to a halt.

    This was most unexpected because, as I said above, there have been no changes to the CSV or the VM's contained within.
    During this space outage I moved some VM's around and re-booted one off the nodes. suddenly the CSV free space jumped to 1.7TB free, which is what I expect but strange as what caused the free space to be used up.

    now a month later the free space is decreasing again (now down from 1.7TB to 970GB). again there has been no changes to the VM's on the CSV

    windirstat puts used pace for the volume at 1.8TB but looking at cluster properties in windows explorer the used space is 2.58TB.

    I do not know why this is.

    Anyone have any ideas other than reboot the nodes to see if this fixes?

    thanks

    Thursday, September 1, 2016 4:24 PM

Answers

  • How many VMs are running on that CSV? 

    Instead of taking reboot, try to migrate owner node of CSV... Online migrate and check if problem is resolved... This might be due to storage related issue not the CSV Volume or cluster issue... I had this issue once and later it was identified the issue was with storage only... make disk offline and online also resolve the issue...

    You can also do one more thing... migrate all VMs from that CSV to another CSV... remove CSV from cluster... format disk and re-add disk as CSV to the cluster...

    Thanks,

    Mark as answer if this helps You!

    Visit: http://www.mdtechskillssolutions.com


    Mayank Dhama

    Thursday, September 1, 2016 4:59 PM
  • Are you using dynamic disks for the VHDs or fixed disks?  You would not have that issue with fixed disks, but dynamic disks do grow over time as space is allocated within them.  They do not automatically shrink when something is deleted.  Moving storage of the VMs around will cause the deleted space within a volume to be freed, thereby freeing the space at the host level.

    . : | : . : | : . tim

    Friday, September 2, 2016 12:10 AM

All replies

  • How many VMs are running on that CSV? 

    Instead of taking reboot, try to migrate owner node of CSV... Online migrate and check if problem is resolved... This might be due to storage related issue not the CSV Volume or cluster issue... I had this issue once and later it was identified the issue was with storage only... make disk offline and online also resolve the issue...

    You can also do one more thing... migrate all VMs from that CSV to another CSV... remove CSV from cluster... format disk and re-add disk as CSV to the cluster...

    Thanks,

    Mark as answer if this helps You!

    Visit: http://www.mdtechskillssolutions.com


    Mayank Dhama

    Thursday, September 1, 2016 4:59 PM
  • Are you using dynamic disks for the VHDs or fixed disks?  You would not have that issue with fixed disks, but dynamic disks do grow over time as space is allocated within them.  They do not automatically shrink when something is deleted.  Moving storage of the VMs around will cause the deleted space within a volume to be freed, thereby freeing the space at the host level.

    . : | : . : | : . tim

    Friday, September 2, 2016 12:10 AM
  • there are around 20 VM's running on the CSV. I did move the CSV owner but problem still remains.

    There are no dynamic VM's running on the CSV

    moving all vm's to the other CSV's will be a bit tricky as they are smaller CSV's running on different disk clusters that run other services.

    There are no major production servers running on the problem CSV so I could shut the VM's down over a weekend and try the CSV offline/online technique.

    The last resort will be the re-format, the VM's are being replicated to a datacentre for DR so that is a viable solution.

    Though it would mean that I would need to start again with the replication of the VM's back to the primary site.

    hoping the CSV offline/online will resolve.

    Friday, September 2, 2016 9:09 AM
  • ok share your result once done :)

    Thanks,

    Mark as answer if this helps You!

    Visit: http://www.mdtechskillssolutions.com


    Mayank Dhama

    Friday, September 2, 2016 4:47 PM
  • I shut down all the VM's over the weekend and off-lined the CSV.

    Unfortunately bringing it back online didn't make a difference. off-lined a few times but no change to the free space :(

    looks like it's going to have to be a CSV format....

    that's going to be a bit more involved and cannot do that immediately.

    Thanks for the suggestions so far.

    Monday, September 5, 2016 8:09 AM
  • Hi Drensta,

    If you got any updates after further tests, welcome to share the information.

    Best Regards,

    Leo


    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Support, contact tnmff@microsoft.com.

    Tuesday, September 6, 2016 5:38 AM
    Moderator
  • Hi Drensta,

    Are there any updates on the issue?

    Best Regards,

    Leo


    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

    Wednesday, September 21, 2016 8:43 AM
    Moderator
  • First things first, be sure to have the latest cluster hotfixes installed -

    Recommended hotfixes and updates for Windows Server
    2012-based failover clusters

    http://support.microsoft.com/kb/2784261

    Recommended hotfixes and updates for Windows Server
    2012 R2-based failover clusters

    http://support.microsoft.com/kb/2920151



    Chuck Timon Senior, Support Escalation Engineer (SEE) Windows 8 Beta Engineer Microsoft Corporation

    Wednesday, September 21, 2016 9:35 PM
  • I to am having the same issue. I am currently running my 4 Hyper-V nodes through Windows Updates and reboots and nothing is seeming to clear up. My EqualLogic SAN is indicating I have over 2TB of free space, but Windows is reporting only 126GB.

    Wondering if anyone else is having the same issue and or has solved this mysterious problem.

    Thanks

    Thursday, September 22, 2016 10:34 PM
  • I am having pretty much the exact problem you are describing. just noticed it today. looking for a solution. 
    Wednesday, September 28, 2016 1:42 PM
  • NOTE: If You all are having SQL Database & Transaction Logs in disk then You can shrink Transaction log & space will be free from disk but still from console view, space will be looks like completely full.

    So first check if any VM with database transaction log is running and how many VMs are running with Dynamic Disk... Go to Hyper-V Console, Click on Inspect Disk and select VM Disk from path to see which disk is utilising more space... Then start further evaluating root cause to identify for more utilisation of disk space.

    Thanks,

    Mark as answer if this helps You!

    Visit: http://www.mdtechskillssolutions.com

     

    Mayank Dhama

    Wednesday, September 28, 2016 3:37 PM
  • Hi All,

    i am facing same kind of problem. Exchange transaction log is running on the VM and the disk type is dynamically expandable. is there any solution to resolve this case.

    Regards,

    Saravanan 

    Wednesday, October 10, 2018 1:35 PM
  • just a reply for anyone that has this issue in the future (2019) - just had the same thing happen in server 2016 after expanding the disk through Fail over cluster manager for one of our VMs.

    on the VM which isnt showing the correct info on the disk

    Open CMD-elevated

    1. diskpart 
    1. diskpart >List Volume
    1. diskpart  >Select Volume X
    1. diskpart >Extend Filesystem

    Friday, September 27, 2019 11:32 PM