Uploaded image for project: 'Lustre'
  1. Lustre
  2. LU-394

LND failure casued by discontiguous KIOV pages

Details

    • Bug
    • Resolution: Fixed
    • Critical
    • Lustre 2.1.0
    • Lustre 2.1.0
    • None
    • 2
    • 4943

    Description

      Cray's gnilnd is running into a hole in kiov list in Lustre 2.1:

      LustreError: 17837:0:(gnilnd_cb.c:594:kgnilnd_setup_phys_buffer()) Can't make payload
      contiguous in I/O VM:page 17, offset 0, nob 6350, kiov_offset 0 kiov_len 2254
      LustreError: 17837:0:(gnilnd_cb.c:1751:kgnilnd_send()) unable to setup buffer: -22

      It used to be that only the first and last page in an IOV were allowed
      to be of a offset + length < PAGE_SIZE.

      It doesn't have this problem with 1.8 client and 2.1 server.

      This problem can be reproduced by "fsx-linux -WR -dn -N 10000 junkfile".

      The osc_brw() is never called and the unfragmented pages logic is not exercised in 2.1

      Attachments

        1. c38.dk.lu394
          6.02 MB
          Wally Wang
        2. nid0037.dk
          722 kB
          Wally Wang

        Activity

          People

            jay Jinshan Xiong (Inactive)
            wang Wally Wang (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            8 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: