Uploaded image for project: 'Lustre'
  1. Lustre
  2. LU-247

Lustre client slow performance on BG/P IONs: unaligned DIRECT_IO

Details

    • Bug
    • Resolution: Fixed
    • Minor
    • None
    • Lustre 2.0.0
    • None
    • 3
    • 18,801
    • 9105

    Description

      Port the fix of bug 18801 to 2.1

      Attachments

        Issue Links

          is related to

          Bug - A problem which impairs or prevents the functions of the product. LU-3192 LBUG:(osc_request.c:1308:osc_brw_prep_request()) ASSERTION( i == 0 || pg->off > pg_prev->off )

          • Major - Major loss of function.
          • Resolved

          Improvement - An improvement or enhancement to an existing feature or task. LU-13805 i/o path: Unaligned direct i/o

          • Major - Major loss of function.
          • Open
          is related to

          Bug - A problem which impairs or prevents the functions of the product. LU-12429 Single client buffered SSF write is slower than O_DIRECT

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Open

          Bug - A problem which impairs or prevents the functions of the product. LU-4664 sync write should consume grant on client

          • Major - Major loss of function.
          • Resolved

          Improvement - An improvement or enhancement to an existing feature or task. LU-4198 Improve IO performance when using DIRECT IO using libaio

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Resolved

          Activity

            [LU-247] Lustre client slow performance on BG/P IONs: unaligned DIRECT_IO
            paf0186 Patrick Farrell added a comment -

            See LU-13805

            paf0186 Patrick Farrell added a comment - See LU-13805
            spitzcor Cory Spitz added a comment -

            If we follow Niu's suggestion to land change #980 and then follow it up, can it land now, or must we wait until after we branch for b2_5?

            spitzcor Cory Spitz added a comment - If we follow Niu's suggestion to land change #980 and then follow it up, can it land now, or must we wait until after we branch for b2_5?
            adilger Andreas Dilger added a comment -

            I don't like the idea of adding new users of the obd_brw() APIs. These are pretty much obsolete on both the client and server, and we should be removing them entirely instead of adding new users.

            adilger Andreas Dilger added a comment - I don't like the idea of adding new users of the obd_brw() APIs. These are pretty much obsolete on both the client and server, and we should be removing them entirely instead of adding new users.
            niu Niu Yawei (Inactive) added a comment -

            The patch in http://review.whamcloud.com/980 uses obd_brw() API in osc_io_dio(), since it requires less code changes and easyer for landing, however, the best way to implement osc_io_dio() should be: introduce a new API which can naturally accept (file offset, bytes, offset in first page, incontiguous pages) and build io requests based on these information directly (brw_page should not be used for direct io, 'pshift' hacking should be eliminated), and osc_io_dio() calls the new API to build io reqeusts and sends them asynchronously.

            The proper implementaion requires more code changes, let's fix it in a follow-up patch.

            niu Niu Yawei (Inactive) added a comment - The patch in http://review.whamcloud.com/980 uses obd_brw() API in osc_io_dio(), since it requires less code changes and easyer for landing, however, the best way to implement osc_io_dio() should be: introduce a new API which can naturally accept (file offset, bytes, offset in first page, incontiguous pages) and build io requests based on these information directly (brw_page should not be used for direct io, 'pshift' hacking should be eliminated), and osc_io_dio() calls the new API to build io reqeusts and sends them asynchronously. The proper implementaion requires more code changes, let's fix it in a follow-up patch.
            niu Niu Yawei (Inactive) added a comment -

            Patch is at http://review.whamcloud.com/980

            niu Niu Yawei (Inactive) added a comment - Patch is at http://review.whamcloud.com/980
            pjones Peter Jones added a comment -

            I think that we will defer this task for now

            pjones Peter Jones added a comment - I think that we will defer this task for now
            green Oleg Drokin added a comment -

            Frankly I am not even sure why do we need radix trees or anything like that for directio or otherwise uncached pages.
            It's a purely one-shot job anyway.

            Lai, your description of patchless io partial page write in clio where unlocked page is marked as clean sounds totally broken.
            Good thing there is no easy way to trigger this mode or we'd have another blocker on our hands.

            green Oleg Drokin added a comment - Frankly I am not even sure why do we need radix trees or anything like that for directio or otherwise uncached pages. It's a purely one-shot job anyway. Lai, your description of patchless io partial page write in clio where unlocked page is marked as clean sounds totally broken. Good thing there is no easy way to trigger this mode or we'd have another blocker on our hands.

            People

              niu Niu Yawei (Inactive)
              niu Niu Yawei (Inactive)
              Votes:
              0 Vote for this issue
              Watchers:
              10 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: