Results 1 to 5 of 5

Thread: Chunk issue

  1. #1
    Join Date
    Sep 2006
    Posts
    6

    Unanswered: Chunk issue

    few months ago two chunks were created in informix,
    but instead of linking raw device file name to the chunk the whole disk
    was linked as a chunk (attached is the screen shot).

    the problem now is that informix has wriiten data to the chunk and we
    are not able to drop and create that chunk again.
    is there a way we can retreive data from that chunk and move it to a
    newly created chunk.
    Attached Thumbnails Attached Thumbnails informix.bmp  

  2. #2
    Join Date
    Aug 2006
    Location
    The Netherlands
    Posts
    248
    Migrate the database(s), with dbexport - dbimport?

  3. #3
    Join Date
    Sep 2006
    Posts
    6
    you mean we should export the full db (around 30 GB) and import it.
    pls tell us the proper syntax to do it.
    but how do we retreive data which is in chunk21 and occupying the whole disk even though it is using only 2 GB.
    we want to free up the disk which was by mistake linked as chunk21
    ie ln -s chunk21 /dev/rdsk/c4t8d0

  4. #4
    Join Date
    Aug 2006
    Location
    The Netherlands
    Posts
    248
    Hi,

    A dbexport and dbimport may be a little awkward to do in this case. I presume none of the chunks belong to the rootdbs? What you could do with dbexport and import is to load only the selected tables from the specific chunks from the unload files dbexport creates per table. You can recognise the tables per chunk through the sysmaster database.

    An easier way probably is to create a new logical volume on another disk and remove the symbolic link the specific chunk refers to and create the same link for the new logical volume. Sounds a little tricky and I certainly wouldn't do it without an actual level 0 archive at hand. But I had to do this recently with mirrored primary chunks and the "relinking" to the mirror chunks went absolutely fine, even in on-line mode...

    A third way I can think of is to produce an ontape level 0 archive, drop the chunks, create new one's and restore only the dbspace(s) the specific chunks are assigned to. Don't know if it will work with smaller target chunks than the archived one's but from IDS 9 down chunkspace past 2 GB can't be adressed by the engine anyway so it doesn't necessarily have to be an issue.

    Anyway I wish you good luck with the job.

    Regards

  5. #5
    Join Date
    Sep 2006
    Posts
    6
    We have dropped and recreated some tables, as of now users are able to work, but when we do ontape to take a level-0 backup,
    we are getting the errors as attached.
    the log also contains other errors which may help u guide us.
    Attached Files Attached Files

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •