none
Protection Group Best Practices and Setup RRS feed

  • Question

  • Wondering where I can find these. I have a lot of servers that are being protected by DPM 2010 and it seems to freeze when there are a lot of jobs running at the same time. I have plenty of bandwidth and the local resources dont seem to be a problem. Whenever this happens I have to cancel most of what's running then the remainder of the jobs start to finish...

    I wanted to find out how the protection groups should be set up with a lot of different types of servers for best performance. I have tried spacing out the jobs far enough from one another but it hasn't seemed to help much... When I run just a few jobs together I never have a problem... Any ideas?

    Thursday, August 19, 2010 5:19 PM

Answers

All replies

  •  

    Hi,

    If you tried configuring the PG to have backups start at staggered times and that didn't help, it sounds to me like a general resource issue on the DPM server. 

    1) Managing Performance http://technet.microsoft.com/en-us/library/ff399471.aspx

    2) Check page file settings http://technet.microsoft.com/en-us/library/ff399244.aspx

    DPM requires a pagefile size that is 0.2 percent the size of all recovery point volumes combined, in addition to the recommended size (generally, 1.5 times the amount of RAM on the computer). For example, if the recovery point volumes on a DPM server total 3 TB, you should increase the pagefile size by 6 GB.

    3) Improving Performance http://technet.microsoft.com/en-us/library/ff399571.aspx

    Also - I would download the whitepapers and reviewing them:  http://technet.microsoft.com/en-us/library/ff833813.aspx

     

     


    Regards, Mike J [MSFT] This posting is provided "AS IS" with no warranties, and confers no rights.
    Thursday, August 19, 2010 9:08 PM
    Moderator
  • Thanks. I

    ll read through all that and see what I can come up with. The times I have most of the problems are when a tape backup and recovery points are all being made at the same time in different protection groups. Synchronizations are usually quick. I'll report back after reading up and making some changes. Hopefully this helps...

     

     

    Thursday, August 19, 2010 9:12 PM
  • Hello,

    DPM will only be as fast is it's slowest bottleneck. For example:
    Min. bandwidth for DPM across the WAN is 512 Kb. If you are backing up 6 TB of data from Florida to California over a 512Kb link, then yes synchs will take a long time and may never finish as the synch may start again before it finished from the first time.

    Staggering the PGs is probably one of the best things you can do to make DPM run efficiently. The links above provided by Mike are an excellent resource.

     

    Thanks Shane 

    Friday, August 20, 2010 2:35 PM
  • Everything here is local on a Gig network. I have a 2GB connection between the server and SAN that is used as the storage.

    I staggered the Sync's hoping it would help. I will be able to report back in a few days to see if that really helped.

    Friday, August 20, 2010 2:38 PM
  • Hello,

    DPM will only be as fast is it's slowest bottleneck. For example:
    Min. bandwidth for DPM across the WAN is 512 Kb. If you are backing up 6 TB of data from Florida to California over a 512Kb link, then yes synchs will take a long time and may never finish as the synch may start again before it finished from the first time.

    Staggering the PGs is probably one of the best things you can do to make DPM run efficiently. The links above provided by Mike are an excellent resource.

     

    Thanks Shane 


    This was not the issue as everything is local on a 2GB connection.

    The solution was to simply stagger the Sync's and make sure there weren't too many jobs running all at one time. Once I figured everything out it seems to be working OK. Running backup to tape and recovery point jobs seem to have the most effect on performance.

    Monday, September 13, 2010 2:04 PM