Uploaded image for project: 'Lustre'
  1. Lustre
  2. LU-4495

client evicted on parallel append write to the shared file.

Details

    • Bug
    • Resolution: Duplicate
    • Critical
    • None
    • Lustre 2.4.0, Lustre 2.5.0, Lustre 2.6.0
    • system with 8 sot's.
    • 3
    • 12299

    Description

      client sometimes evicted with simple workload.

      Attachments

        1. 1590-ptlrpcd
          1.93 MB
        2. 1673-cp_ast
          18 kB
        3. 1884-bl_ast
          4 kB
        4. 2485-main
          1.23 MB
        5. 2487-main
          1.66 MB
        6. mpi_log.c
          0.8 kB
        7. patch
          2 kB

        Issue Links

          is duplicated by

          Bug - A problem which impairs or prevents the functions of the product. LU-4381 clio deadlock from truncate

          • Major - Major loss of function.
          • Resolved

          Activity

            [LU-4495] client evicted on parallel append write to the shared file.
            jlevi Jodi Levi (Inactive) added a comment -

            Reopening to close as duplicate as indicated in the comments.

            jlevi Jodi Levi (Inactive) added a comment - Reopening to close as duplicate as indicated in the comments.
            jay Jinshan Xiong (Inactive) added a comment -

            duplicate of LU-4381.

            jay Jinshan Xiong (Inactive) added a comment - duplicate of LU-4381 .
            paf Patrick Farrell (Inactive) added a comment -

            This is a duplicate of:
            https://jira.hpdd.intel.com/browse/LU-4381

            paf Patrick Farrell (Inactive) added a comment - This is a duplicate of: https://jira.hpdd.intel.com/browse/LU-4381
            paf Patrick Farrell (Inactive) added a comment -

            Jinshan:

            Actually, Alexey's original patch (commenting out code) removes exactly the same code as your patch:

            LU-4381 lov: to not hold sub locks at initialization

            Otherwise, it will cause deadlock because it essentially holds
            some sub locks and then to request others in an arbitrary order.

            Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
            Change-Id: I00d51677aa94ec41856402fd8a027e09355786ee
            Reviewed-on: http://review.whamcloud.com/9152
            Tested-by: Jenkins
            Tested-by: Maloo <hpdd-maloo@intel.com>
            Reviewed-by: Lai Siyao <lai.siyao@intel.com>
            Reviewed-by: Bobi Jam <bobijam@gmail.com>
            Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>

            Which has been landed to master.
            So, nevermind! This bug can be closed.

            paf Patrick Farrell (Inactive) added a comment - Jinshan: Actually, Alexey's original patch (commenting out code) removes exactly the same code as your patch: LU-4381 lov: to not hold sub locks at initialization Otherwise, it will cause deadlock because it essentially holds some sub locks and then to request others in an arbitrary order. Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com> Change-Id: I00d51677aa94ec41856402fd8a027e09355786ee Reviewed-on: http://review.whamcloud.com/9152 Tested-by: Jenkins Tested-by: Maloo <hpdd-maloo@intel.com> Reviewed-by: Lai Siyao <lai.siyao@intel.com> Reviewed-by: Bobi Jam <bobijam@gmail.com> Reviewed-by: Oleg Drokin <oleg.drokin@intel.com> Which has been landed to master. So, nevermind! This bug can be closed.
            pichong Gregoire Pichon added a comment -

            Hello,

            I would like to mention that, on a lustre 2.4.2 version we were seeing client evictions when running a testcase which makes tasks append write to a single shared file.

            After applying patch http://review.whamcloud.com/9876 LU-4558 "clio: Solve a race in cl_lock_put", the client evictions did not occur anymore.

            pichong Gregoire Pichon added a comment - Hello, I would like to mention that, on a lustre 2.4.2 version we were seeing client evictions when running a testcase which makes tasks append write to a single shared file. After applying patch http://review.whamcloud.com/9876 LU-4558 "clio: Solve a race in cl_lock_put", the client evictions did not occur anymore.
            jay Jinshan Xiong (Inactive) added a comment -

            Take it easy, Patrick. To be honest, I didn't look into the patch v2 and v3.

            Anyway, can you please create a patch to comment the code out, as I mentioned in patch v1?

            jay Jinshan Xiong (Inactive) added a comment - Take it easy, Patrick. To be honest, I didn't look into the patch v2 and v3. Anyway, can you please create a patch to comment the code out, as I mentioned in patch v1?

            People

              jay Jinshan Xiong (Inactive)
              shadow Alexey Lyashkov
              Votes:
              0 Vote for this issue
              Watchers:
              14 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: