Results 1 to 5 of 5

Thread: Restoring fails with amrecover with "

  1. #1
    Join Date
    Dec 2011
    Posts
    5

    Default Restoring fails with amrecover with "

    Hello all,
    I have been trying to do a recovery with amrecover, but without success. I have amanda 3.3.0 running on a Debian 6.

    Every time I try to restore files (backup works) it ends with an error "Got no header and data from server, check in amidxtaped.*.debug and amandad.*.debug files on server".

    (I recently applied the fix from "Amrecover 'Always' segfaults - how to debug this ?? Is it related toamanda-client ??")

    But now there is the next problem. I have added the amanda.conf, the amidxtaped.20111223234436.debug and amandad.20111223234235.debug file.

    The only thing that I recognize is in the amidxtaped file "amidxtaped: Insecure dependency in unlink while running setgid at /usr/local/share/perl/5.10.1/Amanda/Changer/disk.pm line 468."
    Attached Files Attached Files

  2. #2
    Join Date
    Mar 2011
    Posts
    9

    Default

    It might be worth looking at my "amrecover issues: partial solution" thread. The need for that solution is easy to identify!

  3. #3
    Join Date
    Dec 2011
    Posts
    5

    Default

    Don't know what happened. amrecover works fine now. I didn't do anything except resetting the environment (the third time) and reducing the amount of data to be backed up. In parallel I had the amidxtaped running (like in post "amrecover issues: partial solution" suggested). In Debian it is under /usr/libexec/amanda

  4. #4
    Join Date
    Feb 2012
    Posts
    1

    Default

    Quote Originally Posted by ets View Post
    Don't know what happened. amrecover works fine now. I didn't do anything except resetting the environment (the third time) and reducing the amount of data to be backed up. In parallel I had the amidxtaped running (like in post "amrecover issues: partial solution" suggested). In Debian it is under /usr/libexec/amanda
    I had the exact same problem... and it was due to killing an amrecover session midway. Seems it left some symlinks and state files in a "running" state, and amidxtaped can't recover because of the setgid issue that perl is trapping. I'm guessing it's not been a huge problem for anyone, since the links and state clean up after themselves normally.

    I cleared the hashes in the state file, and removed all the data links. Probably not the "right" way to clean it up, but it got me working again.

    Hopefully the core team will track the setgid issue down, if they haven't already. But in case they don't, maybe this info will be helpful to others, like me, who found this was the only site/thread discussing the problem.

  5. #5
    Join Date
    Jun 2013
    Posts
    6

    Default

    I'm having the same problem. Could you elaborate on where these "hashes" and "state files" are located? I'd like to try this fix too.

    [UPDATE] If I run the amrecover using strace -f amrecover 2>/tmp/dbg and do the recovery it works fine....I'm running the amrecover
    as root.

    [UPDATE 2] OK, after running with the strace as above this cleared whatever was blocking it from working. So problem is now fixed.
    thx.
    Last edited by mackendw; June 29th, 2013 at 12:59 PM.

Posting Permissions

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