1. Forum Rules (PLEASE CLICK HERE TO READ BEFORE POSTING) Click the link to access ADSM.ORG Acceptable Use Policy and forum rules which should be observed when using this website. Violators may be banned from this website. This message will disappear after you have made at least 12 posts. Thank you for your cooperation.

Newbie needs advice..

Discussion in 'TSM Operation' started by sammiam, Oct 12, 2012.

  1. sammiam

    sammiam New Member

    Joined:
    Oct 2, 2012
    Messages:
    9
    Likes Received:
    0
    Occupation:
    Lead Unix Admin
    Location:
    St. Louis, MO
    Unfortunately, I've inherited the admin of a somewhat sick TSM system... Here's what I've gleaned so far:
    1) maintenance that should be performed is:
    DB Backup
    Expirations
    Disk Storage Pool Backups
    Tape Storage Pool Backups
    Another DB Backup
    Migrations
    Reclamations

    My question... of the above processes, which can run concurrent with each other, and which ones cannot. Am I running them in the correct order? I also need to add 2 additional steps, and they are the move drm, one for getting tapes sent back to us from our offsite storage, and one for pulling local tapes to be sent offsite. Where should I stick these 2 processes.. Obviously, I would think that the DB Backup should be run when the system is quiesed. The Expirations process seems to vary from 1 hour to 3 hours to complete, and re Reclamations have never run to completion, (which is a cause of concern because we were to the point of having no scratch tapes at all, and every thing was going down hill very quickly.)..

    I was able to let the reclamation for our offsite storage pool run for 2 and 1/2 days, and managed to get back 13 scratch tapes. I also ran across a script that would allow me to see tapes that had a very low % of usage, and was able to claim some scratch tapes back from that...

    Honestly, I'm starting to thing that our server is running out of cpu cycles, as it does not appear to be able to get the "daily maintenance" completed to the satisfaction of being able to keep my scratch tape inventory high enough to keep from running out of scratch tapes.

    Initially the expiration and reclamation jobs were set with durations that were dynamically calculated based on the time of day, and knowing when client backups would start kicking off. After talking with some folks, it was pretty much determined that bacause these processes were being terminated early, that they were never really able to reclaim any scratch tapes...

    Thanks every so much for understanding the plight of a newbie...
     
  2.  
  3. moon-buddy

    moon-buddy Moderator

    Joined:
    Aug 24, 2005
    Messages:
    6,150
    Likes Received:
    275
    Occupation:
    Electronics Engineer, Security Professional
    Location:
    Somewhere in the US
    Please see my PM
     
  4. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    The following is the order of each steps I take every day:

    - Backup the primary disk pools to the copy tape pool
    - Backup the primary tape pools to the copy tape pool (incase the disk pool filled and put data in the next pool)
    - Backup the TSM DB to tape
    - Move tapes to the I/O (move drmedia * wherestate=mountable tostate=notmountable)
    - Prepare the recovery plan file and mail them to myself
    - Copy data to the active pools
    - Migrate data from the disk pool to the primary tape pool
    - search for expired data in the DR to reclaim the tapes (EXPIRE INVENTORY)
    - Reclaim storage from expired pools (RECLAIM STGPOOL)
    - Audit Primary Active Disk Pool for errors (only on weekends)

    I hope this helps. I can always send you the bash script if it helps but it's still work in progress.
     

Share This Page