DPM Storage pool has over flowed - recovery point errors prevent me from backing up to disc/tape RRS feed

  • Question

  • Hello,

    I'm not very knowledgable on the Subject of DPM. I have managed to get it working properly til now. I really need some help. And I have not seen a posting here that describes my problem.

    My problem is DPM has filled up the Initial disc that was allocated for it to store data to. Some data will not be allowed to backup any further recovery points. this is preventing my wekkly Tape backup job from running. I back up roughly 4.5 TB of data to tape weekly. The data is a mix of SQL DBs, VMs, & files. all the data goes to Disc - retention range 14 days, and then tape retention range 3 weeks.

    I have a DPM 2010 server/ server 2008 R2 (updates current) with an MD3200i Dell PowerVault SAN attached as a storage pool. Originally, the SAN allocated 2TB on the DPM server. Once testing was complete, the 2 TB was expanded to 9TB. I had to delete all existing protected data from the 2TB original partition, then take the SAN disc offline, then take the SAN disc online before DPM would begin writing to the full 9 TB storage pool. Before I did this it did not recognize any additional space beyond the original 2 TB.

    So, I made it over that event and eventually got DPM to function as advertized. It has been going rather well for the past 4 months until now - the Replica and recovery point volumes are all set to auto-grow, however, they are not auto growing at a rate that is reasonable. for example there is a volume that has 7 GB of data in it, but the volume is 360 GB capacity.

    I have added a 2nd disc to the storage pool another 4 TB - DPM has created 5 dynamic volumes on this disc, but the other 59 volumes remail on the other disc (9TB)

    I have found links for the DPM storage calculator spreadsheet. - it is not helpful. I know how much data storage I need for each protected computer. I see there is a way to shrink the recovery volumes. But I would like to shrink all volumes to a size that is relative to the data that is being backed up.

    I have VMs with fixed disc sizes - I see no reason to have a replica volume that is 1.5 times the Fixed size of the VM data it will never grow into the additional allocated space.

    I also back up regular files in a protected Volume and SQL Databases. I would love for DPM to continue managing the volume growth/storage allocation, but its not doing a good job now - I guess I don't understand how to configure the disc allocations correctly. The defaults have not worked for me.

    Tuesday, November 15, 2011 6:12 PM

All replies

  • Hi,

    I don't see how autogrow could grow a volume with 7GB protected data to be 360GB.   Autogrow grows in 10GB or 25% of volume increments whichever is larger.

    Look for application log events 33143 from source msdpm to see all successful autogrow activity.

    Log Name:      Application
    Source:        MSDPM
    Date:          MM/DD/YY HH:MM:SS AM
    Event ID:      33143
    Task Category: None
    Level:         Information
    Keywords:      Classic
    User:          N/A
    Description:   The volume for load5 has grown from 1.62 GB to 11.62 GB.


    Regards, Mike J. [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.
    Tuesday, November 15, 2011 7:34 PM
  • Mike,

    Thanks for your response. I'm not sure how to proceed from here. Can you answer one more question - I have one VM that is being protected using DPM 2010 the fixed size of the the VHD is 446 GB the Replica volume is 669 GB.

    Based on the following formula my recovery point volume is:

    To calculate the Recovery Point volume:

    For Virtual server data:

    (Data source size x retention range in days x 0.02) + 1600MB

    (446 * 14 * 0.02) + 1.563 = 126.44 GB    note: (1600 MB)/1024 = 1.563 GB

    OK so the calculation for the recovery point volume is 126.00 GB - According to disk Management, this 126 GB volume has 3.85 GB of data in it.

    I'm not clear on why DPM is wasting so much disk space. My VHD is not ever going to get bigger than 446 GB, but DPM has to add 222 GB to the replica volume ????? The recovery point volume is calculated based on Microsoft default space allocations and is wasting 120 GB of space ????

    Can you help me get my 340 GB of space back?

    This same behaviour is causing my whole storage pool to fill up to capacity - My protected data sources have no more space and when it comes time to run a recovery point job or back up the data to tape the jobs are all failing.


    Tuesday, November 22, 2011 10:35 PM
  • Hi,

    If you are using fixed size VHD's, then you need to manually change the default replica size to be closer to the fixed VHD's file sizes.   Recovery point volume size calculation is based on an anticipated  2% data change inside the VHD per day, if the data churn rate is much less, then again, you can reduce the recovery point volume size at the time of protection.

    Wasted recovery point space can be regained by shrinking the recovery point volume using the modify disk allocation wizard. To regain the wasted replica volume space, you need to stop protection, delete the replica (which deletes all recovery points) and reprotect it using the manual replica size.  

    Regards, Mike J. [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.
    Monday, November 28, 2011 2:24 AM