none
Space allocation for one protected Member too much

    Question

  • Hi all

    We have a situation where we are low on space regularly on the DPM server, and appears it may be due to one of the protected sources having too much allocated

    My understanding is that if left to automatic allocation, DPM will allocate twice as much, would i be able to remove only the one member and start again (this member is part of protection Group with other members which don't seem to have double allocated) or would i have to start from scratch

    At the moment as a work around, i am manually removing Recovery Points via Power Shell, which is a bit time consuming

    Any advice appreciated

    DPM version 2012 R2

    Friday, November 16, 2018 11:25 AM

All replies

  • Hello!

    If you have auto-grow checked, and when DPM needs to grow the volume it increases it (assuming sufficient space is available on disk) by 25%.

    Did you try shrinking the recovery point volume by going to Disk allocation?

    You can also modify the disk allocation of a datasource by using the Set-DPMDatasourceDiskAllocation PowerShell cmdlet.

    Best regards,
    Leon


    Blog: https://thesystemcenterblog.com LinkedIn:

    Friday, November 16, 2018 11:50 AM
  • Hi

    Thank you for reply

    Auto grow is enabled, and i can shrink the RP volume via the GUI (via power shell if it wont let me)

    I was hoping to re-allocate this VM space as below, to free some space, i believe you cannot manually reduce the replica Volume, although you can increase

    as below there is a fair bit unused

    cheers



    • Edited by Judge999 Friday, November 16, 2018 2:55 PM
    Friday, November 16, 2018 2:49 PM
  • I believe you would in this case have to your VM from the current protection group and create a new one with auto-grow disabled.



    Blog: https://thesystemcenterblog.com LinkedIn:

    • Proposed as answer by Leon Laude Wednesday, December 19, 2018 2:21 PM
    Friday, November 16, 2018 3:22 PM
  • Hi,

    Just checking in to see if you have any update on this?

    Best regards,
    Leon


    Blog: https://thesystemcenterblog.com LinkedIn:

    Monday, December 10, 2018 7:40 PM
  • Hi,

    Just checking in to see if you have any update on this?

    Best regards,
    Leon


    Blog: https://thesystemcenterblog.com LinkedIn:

    Sorry for late reply

    We created a new PG with auto grow disabled, although we have found we have had to keep increasing RP volume manually, but with concern we will be out of space again

    This new Group contains one VM with SQL, and this may be the reason the space is being quickly used.

    I don't fully understand how Recovery Points are calculated, but read that they maybe more with SQL

    Also we set Retention to only 2 to disk (not alot) , but always says there are 3, is this by design or do i need to change to 1

    Luckily they have plenty of Azure online storage, hence setting low retention to Disk

    Below image shows 2 retention, but a colleague may of already removed one, seems not enough room for a third

    Is there a way of calculating what size a Recovery point will be, is it different for different data types

    Status:    OK
    Replica path:    Click to view details
    Latest recovery point:    09/04/2019 18:01:25
    Oldest recovery point:    08/04/2019 18:02:17
    Total recovery points:    2
    Disk allocation:    Replica volume: 936.05 GB allocated, 624.95 GB used | Recovery point volume: 520.00 GB allocated, 399.59 GB used


    Thursday, April 11, 2019 10:24 AM
  • Unfortunately I cannot see the screenshot you provided, try uploading it again.

    SQL Server workloads will usually grow more than other workloads yes.

    It is difficult to calculate how much a recovery point will be, you could get some information by creating a Disk Utilization report within the Reports pane in DPM, then select group by protection group.


    Blog: https://thesystemcenterblog.com LinkedIn:

    Thursday, April 11, 2019 11:24 AM
  • Hi Sorry image would not paste in, but i just copied text so is in previous reply at bottom

    I will look in to reporting section you mentioned

    I think it's all down to when these DPM servers were built, they were not aware of space issues so don't prepare properly

    We have another DPM server with low disk space on C: where the Azure Scratch folder is located, so occasionally there is not enough space, and online job just sits there for days 0mb

    We had to cancel job, detach VHD, then delete these left over VHD's

    Thanks again

    Thursday, April 11, 2019 1:20 PM