Project online again

Submitted on 16 Jan 2015 9:55:24 UTC

Hi everyone,

Albert@Home is online again.

FYI, we're going to work on a few scheduler improvements over the next days. In general you shouldn't be affected by that but a few scheduler (work) requests might not work temporarily.

Happy crunching,
Oliver

Comments

Oliver Behnke
Oliver Behnke
Moderator
Administrator
Joined: 4 Sep 07
Posts: 320
Credit: 8545955
RAC: 0

We briefly stopped the

16 Jan 2015 12:10:38 UTC
Message 80498
(moderation:
    )

    We briefly stopped the project for a database upgrade...

    Stay tuned,
    Oliver 

    Oliver Behnke
    Oliver Behnke
    Moderator
    Administrator
    Joined: 4 Sep 07
    Posts: 320
    Credit: 8545955
    RAC: 0

    Ok, upgrade done. It took way

    16 Jan 2015 15:20:59 UTC
    Message 80499
    (moderation:
      )

      Ok, upgrade done. It took way longer than expected due to a very low filesystem. We're not yet done with everything planned but we'll leave it at that over the weekend. We're going to carry out the remaining work next week, with the project staying online.

      Best,
      Oliver 

      Viking69
      Joined: 2 Mar 05
      Posts: 19
      Credit: 2024604
      RAC: 0

      At 1:30 AM PST I am getting

      17 Jan 2015 9:36:29 UTC
      Message 80500
      (moderation:
        )

        At 1:30 AM PST I am getting the notice that the project is temporarily offline. So my 2 WU's are not reporting.

        Oliver Behnke
        Oliver Behnke
        Moderator
        Administrator
        Joined: 4 Sep 07
        Posts: 320
        Credit: 8545955
        RAC: 0

        Sorry, the outage over the

        19 Jan 2015 9:05:21 UTC
        Message 80501
        (moderation:
          )

          Sorry, the outage over the weekend was due to a misconfiguration that stopped the project automatically. That's resolved now and the project is up again.

          Please note that we might shut down the scheduler briefly from time to time this week to carry out the pending database work. However, those downtime shouldn't take longer than 2-3 hours.

          Cheers,
          Oliver 

          Oliver Behnke
          Oliver Behnke
          Moderator
          Administrator
          Joined: 4 Sep 07
          Posts: 320
          Credit: 8545955
          RAC: 0

          Most scheduler/database

          19 Jan 2015 13:35:06 UTC
          Message 80502
          (moderation:
            )

            Most scheduler/database maintenance done...

            Viking69
            Joined: 2 Mar 05
            Posts: 19
            Credit: 2024604
            RAC: 0

            MY 2 completed WU's reported

            19 Jan 2015 15:10:18 UTC
            Message 80503
            (moderation:
              )

              MY 2 completed WU's reported but the next 2 had this:

              1/19/2015 7:07:17 AM | Albert@Home | Output file PB0073_01461_46_3_0 for task PB0073_01461_46_3 absent

              Oliver Behnke
              Oliver Behnke
              Moderator
              Administrator
              Joined: 4 Sep 07
              Posts: 320
              Credit: 8545955
              RAC: 0

              I'd need to know the exact

              20 Jan 2015 10:08:00 UTC
              Message 80507
              (moderation:
                )

                I'd need to know the exact host and/or tasks that failed in order to look into it but looking at the publicly available details it's clear that you host can't run any OpenCL tasks, only CUDA tasks succeed so it's clear that there're no output files. You should either verify/fix your OpenCL installation or opt-out of OpenCL tasks entirely (using the account preferences on this site).

                If this needs further discussion please open a new thread in the appropriate forum section.

                HTH,
                Oliver