exited with zero status but no 'finished' file

Forums › Problems and Bug Reports

Sector
Sector
Joined: 15 Dec 05
Posts: 11
Credit: 68711652
RAC: 39529
8 Sep 2011 1:33:10 UTC
Topic 195949
(moderation:
    )

    Quote:
    06-Sep-2011 20:37:23 [Einstein@Home] Task p2030.20090419.G61.17-01.48.S.b2s0g0.00000.dm_2144_0 exited with zero status but no 'finished' file

    All CUDA units are doing this for a few days now, I've tried resetting the project didn't help.

    "NVIDIA GPU 0: GeForce GT 240 (driver version unknown, CUDA version 4000, compute capability 1.2, 511MB, 257 GFLOPS peak)"

    BOINC: 6.12.34
    GPU: Nvidia 240GT (512mb)
    Nvidia drivers: 270.41.19

    CPU units are completing normally, GPU units used to be working fine, without any changes on my end this started happening.

    06-Sep-2011 21:34:07 [---] NVIDIA GPU 0: GeForce GT 240 (driver version unknown, CUDA version 4000, compute capability 1.2, 511MB, 257 GFLOPS peak)

    I updated the Nvidia drivers to: 275.09.07, no change.

    Does anyone have any ideas?

    Gundolf Jahn
    Gundolf Jahn
    Joined: 1 Mar 05
    Posts: 1079
    Credit: 341280
    RAC: 0

    exited with zero status but no 'finished' file

    8 Sep 2011 8:10:33 UTC
    Message 106550
    (moderation:
      )

      Did you see the error message in the erroneous CUDA tasks?

      [21:09:27][15072][INFO ] Starting data processing...
Error: API mismatch: the NVIDIA kernel module has version 270.41.19,
but this NVIDIA driver component has version 275.09.07.  Please make
sure that the kernel module and all NVIDIA driver components
have the same version.
[21:09:27][15072][ERROR] Couldn't initialize CUDA driver API (error: 100)!
[21:09:27][15072][ERROR] Demodulation failed (error: 1020)!
21:09:27 (15072): called boinc_finish


      Gruß,
      Gundolf

      Computer sind nicht alles im Leben. (Kleiner Scherz)

      Sector
      Sector
      Joined: 15 Dec 05
      Posts: 11
      Credit: 68711652
      RAC: 39529

      Nope, no version mismatch

      8 Sep 2011 9:40:38 UTC
      Message 106551 in response to message 106550
      (moderation:
        )

        Nope, no version mismatch problem. This started before I updated the Nvidia drivers. I did that to see if it would help, it didn't. Note if those didn't match wouldn't have even let me restart X11 after I updated the drivers. Though it can happen if I update the drivers but don't restart X11 and leave the old Nvidia module in memory.

        Jord
        Joined: 26 Jan 05
        Posts: 2952
        Credit: 5779100
        RAC: 0

        Did you try a reboot of the

        8 Sep 2011 9:59:05 UTC
        Message 106552 in response to message 106551
        (moderation:
          )

          Did you try a reboot of the system? The only way to reset anything wrong with (something stuck in the memory of) a GPU (on a videocard) is to reboot the whole computer.

          Gundolf Jahn
          Gundolf Jahn
          Joined: 1 Mar 05
          Posts: 1079
          Credit: 341280
          RAC: 0

          RE: Nope, no version

          8 Sep 2011 10:32:33 UTC
          Message 106553 in response to message 106551
          (moderation:
            )

            Quote:
            Nope, no version mismatch problem.


            That excerpt was from your task 246074743 sent back on 7 Sep 2011 4:38:27 UTC.

            Gruß,
            Gundolf

            Computer sind nicht alles im Leben. (Kleiner Scherz)

            Jord
            Joined: 26 Jan 05
            Posts: 2952
            Credit: 5779100
            RAC: 0

            His tasks on September

            8 Sep 2011 11:04:53 UTC
            Message 106554 in response to message 106553
            (moderation:
              )

              His tasks on September 7th

              Quote:
              [20:25:47][5094][INFO ] Seed for random number generator is -1018690778.
              [20:25:49][5094][ERROR] Error creating CUDA FFT plan (error code: 2)
              [20:25:49][5094][ERROR] Demodulation failed (error: 1011)!
              [20:25:49][5094][WARN ] CUDA memory allocation problem encountered!
              ------> Returning control to BOINC, delaying restart for at least five minutes...
              ------> If this problem persists you should consider aborting this task.
              [20:31:04][10374][INFO ] Application startup - thank you for supporting Einstein@Home!
              [20:31:04][10374][INFO ] Starting data processing...
              [20:31:04][10374][INFO ] CUDA global memory status (initial GPU state, including context):
              ------> Used in total: 426 MB (86 MB free / 512 MB total) -> Used by this application (assuming a single GPU task): 0 MB


              September 6th

              Quote:
              [17:44:29][31634][INFO ] Seed for random number generator is 1109422677.
              [17:44:31][31634][ERROR] Error creating CUDA FFT plan (error code: 2)
              [17:44:31][31634][ERROR] Demodulation failed (error: 1011)!
              [17:44:31][31634][WARN ] CUDA memory allocation problem encountered!
              ------> Returning control to BOINC, delaying restart for at least five minutes...
              ------> If this problem persists you should consider aborting this task.
              [17:44:32][31639][INFO ] Application startup - thank you for supporting Einstein@Home!
              [17:44:32][31639][INFO ] Starting data processing...
              [17:44:32][31639][INFO ] CUDA global memory status (initial GPU state, including context):
              ------> Used in total: 383 MB (129 MB free / 512 MB total) -> Used by this application (assuming a single GPU task): 0 MB


              September 5th

              Quote:
              [18:46:42][11321][INFO ] Seed for random number generator is -1021593951.
              [18:46:44][11321][ERROR] Error allocating power spectrum device memory: 25166336 bytes (error: 2)
              [18:46:44][11321][ERROR] Demodulation failed (error: 1006)!
              [18:46:44][11321][WARN ] CUDA memory allocation problem encountered!
              ------> Returning control to BOINC, delaying restart for at least five minutes...
              ------> If this problem persists you should consider aborting this task.
              [18:46:45][11323][INFO ] Application startup - thank you for supporting Einstein@Home!
              [18:46:45][11323][INFO ] Starting data processing...
              [18:46:45][11323][INFO ] CUDA global memory status (initial GPU state, including context):
              ------> Used in total: 315 MB (197 MB free / 512 MB total) -> Used by this application (assuming a single GPU task): 0 MB


              and September 4th, show a memory problem.

              Quote:
              [13:15:46][2366][INFO ] Seed for random number generator is 1141547449.
              [13:15:48][2366][ERROR] Error allocating power spectrum device memory: 25166336 bytes (error: 2)
              [13:15:48][2366][ERROR] Demodulation failed (error: 1006)!
              [13:15:48][2366][WARN ] CUDA memory allocation problem encountered!
              ------> Returning control to BOINC, delaying restart for at least five minutes...
              ------> If this problem persists you should consider aborting this task.
              [13:15:49][2368][INFO ] Application startup - thank you for supporting Einstein@Home!
              [13:15:49][2368][INFO ] Starting data processing...
              [13:15:49][2368][INFO ] CUDA global memory status (initial GPU state, including context):
              ------> Used in total: 315 MB (197 MB free / 512 MB total) -> Used by this application (assuming a single GPU task): 0 MB
              [13:15:49][2368][INFO ] Using CUDA device #0 "GeForce GT 240" (96 CUDA cores / 385.92 GFLOPS)

              Now, of course you can ignore that and continue to try to run tasks, while hoping for a miracle, but it will only be fixed when there's enough GPU memory free. Which can only reliably be done with a system reboot. It cannot be done in any other way, only with a full power cycle.

              Sector
              Sector
              Joined: 15 Dec 05
              Posts: 11
              Credit: 68711652
              RAC: 39529

              RE: RE: Nope, no version

              8 Sep 2011 12:53:36 UTC
              Message 106555 in response to message 106553
              (moderation:
                )

                Quote:
                Quote:
                Nope, no version mismatch problem.

                That excerpt was from your task 246074743 sent back on 7 Sep 2011 4:38:27 UTC.

                Gruß,
                Gundolf

                Right, after the driver update but before I did a reboot. Wasn't in there that long, just left BOINC running while I updated a few other things.

                And problem started before then.

                Note the CUDA units seem to be completing normally now, without changing anything on this end.

                Yes the system has been restarted since the problem began.

                Jord
                Joined: 26 Jan 05
                Posts: 2952
                Credit: 5779100
                RAC: 0

                RE: Yes the system has been

                8 Sep 2011 15:43:08 UTC
                Message 106556 in response to message 106555
                (moderation:
                  )

                  Quote:
                  Yes the system has been restarted since the problem began.


                  Which is what fixed it. Next time you see loads of errors, don't ask, just reboot first. Chances are high that'll instantly fix the problem.

                  Sector
                  Sector
                  Joined: 15 Dec 05
                  Posts: 11
                  Credit: 68711652
                  RAC: 39529

                  RE: RE: Yes the system

                  8 Sep 2011 19:39:28 UTC
                  Message 106557 in response to message 106556
                  (moderation:
                    )

                    Quote:
                    Quote:
                    Yes the system has been restarted since the problem began.

                    Which is what fixed it. Next time you see loads of errors, don't ask, just reboot first. Chances are high that'll instantly fix the problem.

                    Was still doing it for awhile after the reboot.

                    Sector
                    Sector
                    Joined: 15 Dec 05
                    Posts: 11
                    Credit: 68711652
                    RAC: 39529

                    It started doing it again.

                    9 Sep 2011 11:02:10 UTC
                    Message 106558
                    (moderation:
                      )

                      It started doing it again. Just to be clear: A cold boot DID NOT help.

                      Jord
                      Joined: 26 Jan 05
                      Posts: 2952
                      Credit: 5779100
                      RAC: 0

                      Well, if your videocard has

                      9 Sep 2011 17:18:00 UTC
                      Message 106559 in response to message 106558
                      (moderation:
                        )

                        Well, if your videocard has those errors repeatedly, and it continues immediately after a reboot, there's a good chance it's a problem with the videocard (broken memory, broken capacitors, too much heat, etc.). I don't see any other option but to either replace that videocard, or test with another.

                        Comment viewing options

                        Select your preferred way to display the comments and click "Save settings" to activate your changes.

                        Forums › Problems and Bug Reports