Details

    • Question/Request
    • Resolution: Unresolved
    • Minor
    • None
    • None
    • None
    • 9223372036854775807

    Description

      using a 2.15.63 test client under the new redhat 9.4 kernel, which seems to mostly work i'm getting these in dmesg

       

      operation mds_batch to node failed: rc = -524

      layout.c:2316:__req_capsule_get()) @@@ wrong buffer for field 'batch_update_reply' )1 of 1) in format 'MDS_BATCH', 0 vs. 8 (server)

      interpret:RQU/200/0 rc {}524/{}524 job:'' uid:0 gid:0

       

      i was not seeing this under the 2.15.4 and the last redhat 9.3 kernel

      Attachments

        Issue Links

          Activity

            [LU-17834] wrong buffer for field

            the server side is an HPE ClusterStor unit running 2.12

            mdidomenico Michael Di Domenico added a comment - the server side is an HPE ClusterStor unit running 2.12

            Michael, are you running the 2.15.63 client against a 2.15 or older server? This is a known issue and should be fixed by patch https://review.whamcloud.com/55017 "LU-17783 statahead: disable batch statahead for old server". It would be helpful if you applied that patch and confirmed it solves the problem, and +1 the patch in that case so it can be landed.

            adilger Andreas Dilger added a comment - Michael, are you running the 2.15.63 client against a 2.15 or older server? This is a known issue and should be fixed by patch https://review.whamcloud.com/55017 " LU-17783 statahead: disable batch statahead for old server ". It would be helpful if you applied that patch and confirmed it solves the problem, and +1 the patch in that case so it can be landed.

            People

              wc-triage WC Triage
              mdidomenico Michael Di Domenico
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

                Created:
                Updated: