Details

    • Bug
    • Resolution: Fixed
    • Minor
    • Lustre 2.4.0, Lustre 2.1.5
    • Lustre 2.4.0, Lustre 2.1.5
    • None
    • lustre-b2_1/#50
      server RHEL5-x86_64
      client RHEL5-i686
    • 3
    • 2205

    Description

      It looks like an environment issue,please see the Maloo report for more info,test_45g: https://maloo.whamcloud.com/test_sets/1f338df0-929b-11e1-9e8b-525400d2bfa6

      Attachments

        Issue Links

          Activity

            [LU-1486] stat must fail: command not found
            yujian Jian Yu added a comment -

            Patch was landed on Lustre b2_1 branch.

            yujian Jian Yu added a comment - Patch was landed on Lustre b2_1 branch.
            yujian Jian Yu added a comment -

            Patch for master branch: http://review.whamcloud.com/3826

            The patch also needs to be cherry-picked to Lustre b2_1 branch.

            yujian Jian Yu added a comment - Patch for master branch: http://review.whamcloud.com/3826 The patch also needs to be cherry-picked to Lustre b2_1 branch.
            yujian Jian Yu added a comment -

            Patch for master branch: http://review.whamcloud.com/3826

            yujian Jian Yu added a comment - Patch for master branch: http://review.whamcloud.com/3826
            pjones Peter Jones added a comment -

            Yujian

            I'm not sure who the original author is but can you help fix this bug?

            Peter

            pjones Peter Jones added a comment - Yujian I'm not sure who the original author is but can you help fix this bug? Peter
            chris Chris Gearing (Inactive) added a comment - - edited

            This is a bug;

            test_45g() {
            touch $DIR1/$tfile
            #define OBD_FAIL_ONCE|OBD_FAIL_MDS_PDO_LOCK 0x145
            do_facet $SINGLEMDS lctl set_param fail_loc=0x80000145
            mv $DIR1/$tfile $DIR1/$tfile-2 &
            PID1=$!
            sleep 1
            stat $DIR2/$tfile > /dev/null && "stat must fail"
            check_pdo_conflict $PID1 &&

            { wait $PID1; error "getattr isn't blocked"; }

            rm -r $DIR1/*
            return 0
            }

            stat $DIR2/$tfile > /dev/null && "stat must fail"

            should be

            stat $DIR2/$tfile > /dev/null && error "stat must fail"

            This should be assigned to the author of the original test if they are still with the Lustre community.

            chris Chris Gearing (Inactive) added a comment - - edited This is a bug; test_45g() { touch $DIR1/$tfile #define OBD_FAIL_ONCE|OBD_FAIL_MDS_PDO_LOCK 0x145 do_facet $SINGLEMDS lctl set_param fail_loc=0x80000145 mv $DIR1/$tfile $DIR1/$tfile-2 & PID1=$! sleep 1 stat $DIR2/$tfile > /dev/null && "stat must fail" check_pdo_conflict $PID1 && { wait $PID1; error "getattr isn't blocked"; } rm -r $DIR1/* return 0 } stat $DIR2/$tfile > /dev/null && "stat must fail" should be stat $DIR2/$tfile > /dev/null && error "stat must fail" This should be assigned to the author of the original test if they are still with the Lustre community.

            People

              yujian Jian Yu
              sarah Sarah Liu
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: