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.

test restore from copy pool fails... but why?

Discussion in 'Restore / Recovery Discussion' started by foobar2devnull, Oct 16, 2012.

  1. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    Hi all,

    I tried to do a test restore that simulated the failure of the primary pool and hoped it would switch to a copy pool.

    The following is my current setup.

    Code:
    nodes --> PrimaryDisk OS (PD01) +-> Primary File (PF01)
                                    |
                                    +-> Copy Tape (CT01)
                                    |
                                    +-> Copy Tape (CT02) off-site
    
    I did the following steps:
    * backup the node
    * backup PD01 to CT02,CT01
    * migrate PD01 to PF01

    My test:
    * restoring a file was successful through PF01
    Code:
    Restoring       2,155,191 /home/jdoe/sample.tar.gz --> /tmp/sample.tar.gz [Done]
    
    I then deleted all PF01 volumes and ran the restore again.

    I got the following:
    Code:
    ANS1302E No objects on server match query
    
    I thought it would automatically search for the file on a copy pool when not found in the primary pool. I suspect I did something wrong but what?

    Can you advise?

    Thanks!
     
  2.  
  3. moon-buddy

    moon-buddy Moderator

    Joined:
    Aug 24, 2005
    Messages:
    6,263
    Likes Received:
    282
    Occupation:
    Electronics Engineer, Security Professional
    Location:
    Somewhere in the US
    Deleted or marked destroyed?
     
  4. chad_small

    chad_small Moderator

    Joined:
    Dec 17, 2002
    Messages:
    2,207
    Likes Received:
    44
    Occupation:
    AIX/SAN/TSM
    Location:
    Gilbert, AZ
    Whats the status of the copy volumes? Are they shown as onsite/readonly?
     
  5. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    Sorry for the latre reply, we're in different time zones ;)

    moon-buddy, the answer to your question is "deleted". I ran a "delete vol" on all the volumes in the PF01.

    chad_small, both copy pools are onsite. I have not setup the DRM yet.

    >select STGPOOL_NAME,POOLTYPE,ACCESS from stgpools where STGPOOL_NAME like 'CT0%'

    STGPOOL_NAME POOLTYPE ACCESS
    -------------------------------- -------------------------------- ---------------------------------------------------------------------------------------------------------------------------------
    CT01 COPY READWRITE
    CT02 COPY READWRITE


    Thank you both for your help. I'll keep looking.
     
    Last edited: Oct 17, 2012
  6. Jeroen

    Jeroen Senior Member

    Joined:
    Oct 2, 2002
    Messages:
    356
    Likes Received:
    12
    Occupation:
    AIX + storage administrator
    Location:
    netherlands
    "delete vol XX discarddata=yes" will also delete the data on the copy volumes.
     
    foobar2devnull likes this.
  7. moon-buddy

    moon-buddy Moderator

    Joined:
    Aug 24, 2005
    Messages:
    6,263
    Likes Received:
    282
    Occupation:
    Electronics Engineer, Security Professional
    Location:
    Somewhere in the US
    Yep - as what Jeroen said - deleting the volumes with 'discarddata=yes' deletes all copies associated with that volume. The proper way - and this is simulating a disaster - is to mark the onsite copy pool as destroyed.

    I hope you have the DB backup prior to doing all of these activities. You will need to restore that DB backup to get back all of your data.
     
    Last edited: Oct 25, 2012
    foobar2devnull likes this.
  8. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    Thank you all for your help. The test was made on a new server with my data alone so no loss there but rather a good lesson learned! :)
     
  9. pburwell

    pburwell New Member

    Joined:
    Apr 5, 2013
    Messages:
    14
    Likes Received:
    0
    Location:
    USA, PA
    I have a similar issue except it's a corrupted volume that I am trying to REMOVE from the stgpool it's in. I don't want to delete the data registered on the volume anywhere else, so how do I just get rid of the volume from the stagepool? I have already removed the VOL from the library and yet it remains listed in the stgpool.
    Ideas?
     
  10. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
  11. pburwell

    pburwell New Member

    Joined:
    Apr 5, 2013
    Messages:
    14
    Likes Received:
    0
    Location:
    USA, PA
    I marked it as destroyed, recovered the data from the recovery volumes, checked out the volume from the library and ejected the tape. But now it is still showing in the stgpool. What now?
     
    Last edited: Apr 8, 2013
  12. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    I'm no veteran and I'm not in front of my console but does the CHECKOUT command not put the tape into the I/O and remove the entry from the library list. What command do you run to see it is still listed?

    Code:
    Q LIBVOL <tape>
    Maybe an audit of the library followed by one of TSM might be in order. Any one else have advice on this?
     
  13. pburwell

    pburwell New Member

    Joined:
    Apr 5, 2013
    Messages:
    14
    Likes Received:
    0
    Location:
    USA, PA
    I want to remove the entry from the STGPOOL. It's already removed from the Library:
    But it still shows the volume is IN the Primary Stagepool I am trying to clear:
    But the volume shows Destroyed:

    I cannot seem to get the entry for the tape out, even though it is physically removed from the 3494, because:

    Seems to be a loxadrome!
     
    Last edited: Apr 8, 2013
  14. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    I believe that the order in which you do your steps is important.

    Did you mark the tape as destroyed before or after checking it out?

    Did your 'restore volume' command work? It complains there is still data on the tape so I wonder...

    Maybe you should check it back in, see if a 'q libvol' will work and then check it out again before attempting another 'del vol'
     
  15. pburwell

    pburwell New Member

    Joined:
    Apr 5, 2013
    Messages:
    14
    Likes Received:
    0
    Location:
    USA, PA
  16. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    Ok, so my first question would be, do you have a copy pool?

    1) If not, Step '4' onwards are of no use to you and you should probably put the tape back into the library, mark it 'readonly' and try and move the data to another tape (Steps 1-3).

    2) If you do have a copy pool, it looks like the 'restore volume' command might have failed. If unable to rebuild the volume, you won't be able to delete it as it still holds valid data. Try steps 4 onwards or follow the IBM document I added to my initial post.

    Best of luck.
     
  17. pburwell

    pburwell New Member

    Joined:
    Apr 5, 2013
    Messages:
    14
    Likes Received:
    0
    Location:
    USA, PA
    Yes.
    Tape is corrupt anyway, cannot mv data from it.

    This is apparently what TSM S/W support is trying to establish. I am re-running the "Restore Vol" tasks. Will apprise what the results are... initially it appears TSM is not tracking data location correctly. Hopefully this is NOT the case.
    I moved this data sucessfully tweice now and yet Query Content still "shows" data for this destroyed 3590 volume.

    Thanks. :)

    -Patrick
     
  18. foobar2devnull

    foobar2devnull New Member

    Joined:
    Nov 30, 2010
    Messages:
    97
    Likes Received:
    1
    Occupation:
    System and Backup Administrator
    Location:
    Belgium
    For what it's worth

    I had to recover from a defect tape just yesterday so I thought the output of a successful volume restore might help you identify where it went wrong for you.

    Code:
    04/08/2013 11:51:01      ANR2017I Administrator JDOE issued command: RESTORE                          
                              VOLUME AA0049L5 MAXPROCESS=2  (SESSION: 145076)
    04/08/2013 11:51:12      ANR2114I RESTORE VOLUME: Access mode for volume AA0049L5
                              updated to "destroyed". (SESSION: 145076)
    04/08/2013 11:51:12      ANR0984I Process 324 for RESTORE VOLUME started in the
                              BACKGROUND at 11:51:12 AM. (SESSION: 145076, PROCESS:
                              324)
    04/08/2013 11:51:12      ANR2110I RESTORE VOLUME started as process 324. (SESSION:
                              145076, PROCESS: 324)
    04/08/2013 11:51:12      ANR0984I Process 325 for RESTORE VOLUME started in the
                              BACKGROUND at 11:51:12 AM. (SESSION: 145076, PROCESS:
                              325)
    04/08/2013 11:51:12      ANR2110I RESTORE VOLUME started as process 325. (SESSION:
                              145076, PROCESS: 325)
    
    
    [...]
    
    
    04/08/2013 16:59:04      ANR0986I Process 325 for RESTORE VOLUME running in the
                              BACKGROUND processed 1,126,587 items for a total of
                              1,609,635,922,941 bytes with a completion state of
                              SUCCESS at 04:59:04 PM. (SESSION: 145076, PROCESS: 325)
    04/08/2013 17:46:13      ANR0986I Process 324 for RESTORE VOLUME running in the
                              BACKGROUND processed 3,806,153 items for a total of
                              1,285,128,362,038 bytes with a completion state of
                              SUCCESS at 05:46:13 PM. (SESSION: 145076, PROCESS: 324)
    04/08/2013 17:46:13      ANR1240I Restore of volumes in primary storage pool PT01
                              has ended.  Files Restored: 4932740, Bytes Restored:
                              2894764284979, Unreadable Files: 0, Unreadable Bytes: 0.
                              (SESSION: 145076)
    04/08/2013 17:46:13      ANR2208I Volume AA0049L5 deleted from storage pool PT01.
                              (SESSION: 145076)
    04/08/2013 17:46:13      ANR1341I Scratch volume AA0049L5 has been deleted from
                              storage pool PT01. (SESSION: 145076)
    Best of luck
     
  19. pburwell

    pburwell New Member

    Joined:
    Apr 5, 2013
    Messages:
    14
    Likes Received:
    0
    Location:
    USA, PA
    It's ANOTHER IT Miracle!

    Well now! That worked right this time! :)
    I had complained to a senior TSM engineer and suddenly it worked this time (while the last two times failed).
    It's an IT MIRACLE! (Yeah right)
    Well I gues we are done. Thanks!
     
    Last edited: Apr 10, 2013

Share This Page