1. 28 Dec, 2009 2 commits
  2. 18 Dec, 2009 1 commit
  3. 14 Dec, 2009 3 commits
  4. 25 Sep, 2009 6 commits
  5. 11 Sep, 2009 1 commit
  6. 08 Sep, 2009 1 commit
  7. 07 Sep, 2009 1 commit
    • Iustin Pop's avatar
      Optimise multi-job submit · 009e73d0
      Iustin Pop authored
      
      
      Currently, on multi-job submits we simply iterate over the
      single-job-submit function. This means we grab a new serial, write and
      replicate (and wait for the remote nodes to ack) the serial file, and
      only then create the job file; this is repeated N times, once for each
      job.
      
      Since job identifiers are ‘cheap’, it's simpler to simply grab at the
      start a block of new IDs, write and replicate the serial count file a
      single time, and then proceed with the jobs as before. This is a cheap
      change that reduces I/O and reduces slightly the CPU consumption of the
      master daemon: submit time seems to be cut in half for big batches of
      jobs and the masterd cpu time by (I can't get consistent numbers)
      between 15%-50%.
      
      Note that this doesn't change anything for single-job submits and most
      probably for < 5 job submits either.
      Signed-off-by: default avatarIustin Pop <iustin@google.com>
      Reviewed-by: default avatarMichael Hanselmann <hansmi@google.com>
      009e73d0
  8. 31 Aug, 2009 1 commit
  9. 25 Aug, 2009 1 commit
  10. 21 Aug, 2009 1 commit
  11. 20 Aug, 2009 1 commit
  12. 19 Aug, 2009 1 commit
  13. 17 Aug, 2009 1 commit
  14. 14 Aug, 2009 2 commits
  15. 13 Aug, 2009 1 commit
  16. 12 Aug, 2009 1 commit
  17. 11 Aug, 2009 1 commit
  18. 10 Aug, 2009 6 commits
  19. 05 Aug, 2009 2 commits
    • Iustin Pop's avatar
      export: add meaningful exit code · 084f05a5
      Iustin Pop authored
      
      
      Currently ‘gnt-backup export’ always returns exit code zero, even in the
      face of complete failure during backup (only failure to stop/start the
      instance will cause job failure and thus non-zero exit code). This is
      bad, since one cannot script the backup.
      
      This patch adds some simple results from the LU so that the command line
      script can return good exit code. It will:
        - return zero for full success (snapshot removal errors are ignored
          though)
        - return one for full failure (finalize export failure or all disks
          failure)
        - return two for partial failure (some disks backed up, some not)
      Signed-off-by: default avatarIustin Pop <iustin@google.com>
      Reviewed-by: default avatarMichael Hanselmann <hansmi@google.com>
      084f05a5
    • Iustin Pop's avatar
      Fix detecting of errors in export · a48b08bf
      Iustin Pop authored
      
      
      This should fix issue 61, by explicitely calling bash (which is is now a
      non-explicit dependency) and setting the pipefail command.
      Signed-off-by: default avatarIustin Pop <iustin@google.com>
      Reviewed-by: default avatarMichael Hanselmann <hansmi@google.com>
      a48b08bf
  20. 04 Aug, 2009 6 commits