Uploaded image for project: 'Lustre'
  1. Lustre
  2. LU-6455

EL7 client replay-vbr test_4i: version changed unexpectedly

Details

    • Bug
    • Resolution: Fixed
    • Blocker
    • Lustre 2.8.0, Lustre 2.10.0
    • Lustre 2.8.0
    • None
    • client and server: lustre-master build # 2983 EL7 client
    • 3
    • 9223372036854775807

    Description

      This issue was created by maloo for sarah <sarah@whamcloud.com>

      This issue relates to the following test suite run: https://testing.hpdd.intel.com/test_sets/ce29134a-df8a-11e4-bf2e-5254006e85c2.

      The sub-test test_4i failed with the following error:

      version changed unexpectedly: pre 0x8d00000008, post 0x8d0000000d
      

      test log

      == replay-vbr test 4i: setattr of times does not change versions ===================================== 20:19:13 (1428635953)
      CMD: onyx-42vm3 /usr/sbin/lctl get_param mdd.lustre-MDT0000.atime_diff
      CMD: onyx-42vm3 /usr/sbin/lctl set_param mdd.lustre-MDT0000.atime_diff=0
      mdd.lustre-MDT0000.atime_diff=0
      CMD: onyx-42vm6.onyx.hpdd.intel.com mcreate /mnt/lustre/f4i.replay-vbr
      CMD: onyx-42vm6.onyx.hpdd.intel.com /usr/bin/lfs path2fid /mnt/lustre/f4i.replay-vbr
      CMD: onyx-42vm3 /usr/sbin/lctl --device lustre-MDT0000 getobjversion \"[0x200014450:0x1:0x0]\"
      CMD: onyx-42vm6.onyx.hpdd.intel.com touch /mnt/lustre/f4i.replay-vbr
      CMD: onyx-42vm6.onyx.hpdd.intel.com /usr/bin/lfs path2fid /mnt/lustre/f4i.replay-vbr
      CMD: onyx-42vm3 /usr/sbin/lctl --device lustre-MDT0000 getobjversion \"[0x200014450:0x1:0x0]\"
      CMD: onyx-42vm3 /usr/sbin/lctl set_param mdd.lustre-MDT0000.atime_diff=60
      mdd.lustre-MDT0000.atime_diff=60
       replay-vbr test_4i: @@@@@@ FAIL: version changed unexpectedly: pre 0x8d00000008, post 0x8d0000000d 
        Trace dump:
      

      Attachments

        Issue Links

          Activity

            [LU-6455] EL7 client replay-vbr test_4i: version changed unexpectedly

            Oleg Drokin (oleg.drokin@intel.com) merged in patch https://review.whamcloud.com/21565/
            Subject: LU-6455 tests: Re-enable replay-vbr and replay-single tests
            Project: fs/lustre-release
            Branch: master
            Current Patch Set:
            Commit: 8b98c8a669d8d6c36f8bd767acba5e6f912360b7

            gerrit Gerrit Updater added a comment - Oleg Drokin (oleg.drokin@intel.com) merged in patch https://review.whamcloud.com/21565/ Subject: LU-6455 tests: Re-enable replay-vbr and replay-single tests Project: fs/lustre-release Branch: master Current Patch Set: Commit: 8b98c8a669d8d6c36f8bd767acba5e6f912360b7

            James Nunez (james.a.nunez@intel.com) uploaded a new patch: http://review.whamcloud.com/21565
            Subject: LU-6455 tests: Re-enable replay-vbr 4i-k and 10b
            Project: fs/lustre-release
            Branch: master
            Current Patch Set: 1
            Commit: 19ee53fc9b242e3871b426a28d390f346d9dbb5f

            gerrit Gerrit Updater added a comment - James Nunez (james.a.nunez@intel.com) uploaded a new patch: http://review.whamcloud.com/21565 Subject: LU-6455 tests: Re-enable replay-vbr 4i-k and 10b Project: fs/lustre-release Branch: master Current Patch Set: 1 Commit: 19ee53fc9b242e3871b426a28d390f346d9dbb5f
            yujian Jian Yu added a comment -

            I think yes, James. Could you please push a patch to enable the tests and add test parameters to verify them on el7?

            yujian Jian Yu added a comment - I think yes, James. Could you please push a patch to enable the tests and add test parameters to verify them on el7?

            replay-single test 28 and replay-vbr tests 4i, 4j, 4k and 10b are still on the ALWAYS_EXCEPT (not being run) list for el7. Since all of the patches for this ticket have landed, can we run these tests again on el7?

            jamesanunez James Nunez (Inactive) added a comment - replay-single test 28 and replay-vbr tests 4i, 4j, 4k and 10b are still on the ALWAYS_EXCEPT (not being run) list for el7. Since all of the patches for this ticket have landed, can we run these tests again on el7?

            Oleg Drokin (oleg.drokin@intel.com) merged in patch http://review.whamcloud.com/14928/
            Subject: LU-6455 mdt: disable IMA support
            Project: fs/lustre-release
            Branch: master
            Current Patch Set:
            Commit: 259c1ec0cfb76b2645efa2be6ec7ad48229eb658

            gerrit Gerrit Updater added a comment - Oleg Drokin (oleg.drokin@intel.com) merged in patch http://review.whamcloud.com/14928/ Subject: LU-6455 mdt: disable IMA support Project: fs/lustre-release Branch: master Current Patch Set: Commit: 259c1ec0cfb76b2645efa2be6ec7ad48229eb658
            pjones Peter Jones added a comment -

            This has been done now and so the this test should no longer be failing

            pjones Peter Jones added a comment - This has been done now and so the this test should no longer be failing

            I think the needed rpm is 'redhat-lsb'. should be installed on test nodes. should be installed everywhere.

            bogl Bob Glossman (Inactive) added a comment - I think the needed rpm is 'redhat-lsb'. should be installed on test nodes. should be installed everywhere.
            green Oleg Drokin added a comment -

            So the patch has a flaw assuming lsb_release is present when it's not.

            From the log:

            20:12:25:/usr/lib64/lustre/tests/replay-vbr.sh: line 10: lsb_release: command not found
            
            green Oleg Drokin added a comment - So the patch has a flaw assuming lsb_release is present when it's not. From the log: 20:12:25:/usr/lib64/lustre/tests/replay-vbr.sh: line 10: lsb_release: command not found
            sarah Sarah Liu added a comment -

            still hit this issue on EL7 lustre-master build # 3093 ZFS

            https://testing.hpdd.intel.com/test_sets/19b0e0c4-2917-11e5-b07d-5254006e85c2

            sarah Sarah Liu added a comment - still hit this issue on EL7 lustre-master build # 3093 ZFS https://testing.hpdd.intel.com/test_sets/19b0e0c4-2917-11e5-b07d-5254006e85c2

            The IMA-affected tests have been disabled for RHEL7 only. Unfortunately the fact that the IMA feature sets an extra xattr automatically means that there isn't any way that the replay-vbr.sh tests can verify correct operation. At best we might do something like check if the inode version was increased only once when IMA is enabled, but that is tricky.

            A small improvement would be run the tests anyway and make them error_ignore for RHEL7, so that at least they are being run, but we can address this when the next distro using IMA appears (SLES13 or whatever).

            adilger Andreas Dilger added a comment - The IMA-affected tests have been disabled for RHEL7 only. Unfortunately the fact that the IMA feature sets an extra xattr automatically means that there isn't any way that the replay-vbr.sh tests can verify correct operation. At best we might do something like check if the inode version was increased only once when IMA is enabled, but that is tricky. A small improvement would be run the tests anyway and make them error_ignore for RHEL7, so that at least they are being run, but we can address this when the next distro using IMA appears (SLES13 or whatever).

            People

              hongchao.zhang Hongchao Zhang
              maloo Maloo
              Votes:
              0 Vote for this issue
              Watchers:
              13 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: