Results 1 to 2 of 2
  1. #1
    Join Date
    Nov 2004
    Posts
    2

    Unhappy Unanswered: Please Help: Recover from a corrupted device

    Hi, I'm desparately seeking for someone who could point me to some useful information regarding a disaster recovery problem. Any help is appreciated.

    We are running ASE 11.9.3 (64 bit) on Sun Solaris 7, approx. 20GB of data, 10 raw devices, 2 GB each.

    - one raw device (HD) crashed
    - no working backup exists (due to a completely other disaster, unfortunately)
    - HD was restored in a laboratory (copied) and mounted successfully again
    - master, sybsystemprocs and tempdb are fine

    Problem: The user database causes the recovery process to fail on startup, throwing an "infected by 11" exception (see backtrace below).

    I started ASE in single-user mode, bypassing the recovery and made a dump of the database. It worked fine. Now at least a complete (but still corrupt) backup exists. Maybe it's a small sign of hope.

    Attempts to extract individual tables via bcp failed ("soft" page allocation errors occured, bcp exited). Attempts to run dbcc checkalloc lead to the same exception (below). After studying the Sybase documentation again and again, I have no more ideas.

    Question: Is there any way to get the restored raw device running again? Perhaps some kind of "reformat" without total data loss? Partial data loss would be no problem.

    Any advice is highly appreciated. Thanks in advance for your help.

    Regards,
    Yama


    -------------------------
    This is the error thrown:

    kernel current process (0x30003) infected with 11
    kernel Address 0x00000000802099f4 (getpage__check_timestamps+0x94), siginfo (
    code, address) = (1, 0xffffffff8d8d67fe)
    kernel ************************************
    kernel curdb = 5 pstat = 0x1000 lasterror = 0
    kernel preverror = 0 transtate = 1
    kernel curcmd = 0 program =
    kernel pc: 0x0000000080525c18 pcstkwalk+0x20...
    kernel pc: 0x00000000805008a0 ucbacktrace+0xb0...
    kernel pc: 0x00000000800ac6b8 terminate_process+0xba0 ...
    kernel pc: 0x0000000080512cd4 kisignal+0x94 ...
    kernel pc: 0xffffffff7eda268c _setuid+0x4c ...
    kernel pc: 0x00000000801182cc bufread+0x15c ...
    kernel pc: 0x0000000080206a7c getpage_with_validation+0x2a4 ...
    kernel pc: 0x0000000080177758 pg__readalloc+0x48 ...
    kernel pc: 0x000000008017bafc pg_allocpg_extent_count+0xc ...
    kernel pc: 0x00000000801b3b40 th__count_unreserved_space+0x490 ...
    kernel pc: 0x00000000801b35e8 th_fill_freespaceinfo+0x1d0 ...
    kernel [Handler pc: 0x00000000802a31d8 rec_handle installed by the following function:-]
    kernel pc: 0x00000000802a08f8 recovery+0x9c0 ...
    ....
    Last edited by Yamamoto; 11-30-04 at 12:56.

  2. #2
    Join Date
    Nov 2004
    Posts
    2

    More Information

    I should add that this message occurs frequently during dbcc checkdb:

    Uninitialized logical page '6091264' was read while accessing object '99' in
    database '5'. Please contact Sybase Technical Support.

    The restored HD is ok on the hardware level but obviously has inconsistencies on the data level which were "copied" from the defect disk.

    Yama

Posting Permissions

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