climateprediction.net home page

"Project has no tasks available"


Advanced search

Message boards : Number crunching : "Project has no tasks available"

AuthorMessage
richard schumacher
Send message
Joined: Feb 20 06
Posts: 22
Credit: 2,140,455
RAC: 0
Message 45063 - Posted 12 Oct 2012 13:00:37 UTC

    Since Monday 08 October 2012. Really?
    ____________

    Les Bayliss
    Forum moderator
    Send message
    Joined: Sep 5 04
    Posts: 5348
    Credit: 8,876,229
    RAC: 549
    Message 45064 - Posted 12 Oct 2012 13:25:11 UTC - in response to Message 45063.

      Yes. Really. :)

      For the past 4-5 years, the project has relied on external sources for it's work.
      This is supplied in batches of a few thousand data sets at a time. Then the researchers wait until sufficient models have been completed, so that they can see the results before deciding what area to examine next.

      And there's well over 35,000 computers looking for work, so it doesn't last long when it DOES become available.


      ____________
      Backups: Here

      Bjarke
      Send message
      Joined: Aug 23 06
      Posts: 2
      Credit: 2,307,804
      RAC: 1,877
      Message 45065 - Posted 12 Oct 2012 13:34:46 UTC

        Any idea on the time horizon for new jobs to be available? Perhaps just the order of magnitude eg. days, weeks, months?

        Les Bayliss
        Forum moderator
        Send message
        Joined: Sep 5 04
        Posts: 5348
        Credit: 8,876,229
        RAC: 549
        Message 45066 - Posted 12 Oct 2012 14:10:45 UTC - in response to Message 45065.

          Short answer, No.

          Long answer, the one often used by Sir Humphrey Appleby, in "Yes Minister/Yes Prime Minister:
          "In the fullness of time; When the time is ripe; etc, etc."

          I've forgotten the rest, but it may be in this list of quotes.


          ____________
          Backups: Here

          richard schumacher
          Send message
          Joined: Feb 20 06
          Posts: 22
          Credit: 2,140,455
          RAC: 0
          Message 45073 - Posted 13 Oct 2012 0:19:30 UTC - in response to Message 45064.

            Thanks! Science marches on :_>
            ____________

            Profile JIM
            Send message
            Joined: Dec 31 07
            Posts: 676
            Credit: 3,957,635
            RAC: 2,780
            Message 45075 - Posted 13 Oct 2012 4:21:02 UTC - in response to Message 45073.

              Dear Richard:

              I’ve found that the best way to keep my computers in work is to set the work buffer in the Boinc Manager at 9 days and pray. This lets my machines start begging for work more than a week before I have empty cores. It usually works, but, it also pays be attached to short projects like “Malaria Control” or Rosetta@home. I keep them on “no new tasks” and can enable them any time I have an empty core. That way I can run a few of their WU’s while waiting for CP to get more work. Just remember to change the work buffer setting to about ½ day before opening the flood gates on one of these backup projects. You don’t want nine day worth of WU’s all at once.

              ____________

              3rkko
              Send message
              Joined: Feb 12 08
              Posts: 54
              Credit: 4,247,765
              RAC: 24
              Message 45077 - Posted 13 Oct 2012 20:52:50 UTC - in response to Message 45063.

                Any news on more tasks would be welcome. In the mean time I'll let the various projects at WCG to get more than a fair share of my cores. :)

                BarryAZ
                Send message
                Joined: Jul 13 05
                Posts: 121
                Credit: 11,298,043
                RAC: 73
                Message 45078 - Posted 14 Oct 2012 5:38:09 UTC - in response to Message 45077.

                  My sense is that Les may be unhappy that he has no more responsive answers regarding work availability than he has been able to provide. This is similar to the lack of timing information he has to provide when there are disk storage problems.

                  Seemingly that sort of information is considered confidential by the folks who actually work on the project. Perhaps it is a case of the project folks seeking, by holding project status information back, to address the problem of too many processors for the project to handle by 'guiding people' away from this project to other projects which tend to be more responsive to status and information requests.

                  In this scenario, Les is caught in the middle -- his last post in this thread suggests to me a bit of discomfort with that role.

                  Les Bayliss
                  Forum moderator
                  Send message
                  Joined: Sep 5 04
                  Posts: 5348
                  Credit: 8,876,229
                  RAC: 549
                  Message 45082 - Posted 14 Oct 2012 8:16:02 UTC

                    There is a bit of that, Barry. It seems that no matter how often one of us says that there's no information available on when, someone will still persist in wanting to know.
                    I'm just not going to bother any more, especially as the timing will be obvious by work suddenly becoming available.

                    And the project people themselves don't always know either. They just create data sets from information about years and parameters received from the research groups.
                    The South African group seem to have stopped submitting data, as have the people running the RAPID-RAPIT project.


                    ____________
                    Backups: Here

                    rwremote
                    Send message
                    Joined: Feb 15 11
                    Posts: 6
                    Credit: 1,016,516
                    RAC: 0
                    Message 45083 - Posted 14 Oct 2012 13:44:22 UTC

                      I'm thinking its people not "reading fully" what has been written previously.
                      12th - Have we no tasks ?
                      12th - No we really haven't
                      12th - Any idea when ?
                      12th - Unfortunately no. Due to . . . . [nice long explanation].
                      13th - Any more news . . . . .
                      13th - ahhhhhhhhhhhhhh !!!!!!!!!!!!

                      There are Crunchers and there are Scientists - and then there are the people who have the envious(lol) job of being the buffer-zone/interface between the two.
                      Thankless task if you ask me, but someone has to do it . . . . . or do they ? :)

                      Keep going Les, we do appreciate all the moderators information and feedback.

                      Profile PatrickProject donor
                      Send message
                      Joined: Sep 8 10
                      Posts: 6
                      Credit: 1,052,227
                      RAC: 32
                      Message 45118 - Posted 20 Oct 2012 19:30:23 UTC

                        So I did what I've never done before and added another project, World Community Grid. I'd really like to work on climate change but I don't see any other projects devoted to the subject and who knows when more tasks will arrive at cpdn. Like I said before, cpdn is "set it and forget it."

                        World Community Grid is chugging along and processing tasks very quickly while my remaining cpdn uploads have been stuck in the queue for at least a day or two.

                        There's an IBM logo stuck on the WCG web site so I'm assuming that WCG may have IBM resources behind it.

                        I see that the US National Center for Atmnospheric Research's solution was to acquire the faster supercomputer yet devoted to climate change.

                        http://www.sustainablebusiness.com/index.cfm/go/news.display/id/24193

                        Wonder how that stacks up again a public, distributed computing infrastructure (like cpdn)? The supercomputer is probably more powerful. What's certain though is that with the dedicated supercomputer, the scientists have more control and no bothersome user community to attend to.

                        Profile tullio
                        Send message
                        Joined: Aug 6 04
                        Posts: 183
                        Credit: 300,145
                        RAC: 201
                        Message 45119 - Posted 21 Oct 2012 4:53:55 UTC

                          I am running other 6 BOINC projects on 2 computers: LHC@home, Test4Theory@home, both from CERN, Einstein@home and Albert@home from the Max Planck Gravitational Physics Institute in Hannover, QMC@home from Muenster U. and now transferring to the Bonn Theoretical Chemistry Institute and the daddy of BOINC, SETI@home, so I am never out of work even if some of them are temporarily out of work or transferring, like QMC@home.
                          Tullio
                          ____________

                          Profile Dave Jackson
                          Send message
                          Joined: May 15 09
                          Posts: 811
                          Credit: 632,379
                          RAC: 338
                          Message 45120 - Posted 21 Oct 2012 9:48:42 UTC

                            If you keep a close eye on the server status page, you will see there are tasks but they are being grabbed so quickly that they don't appear or if they do, only very briefly. The number of tasks in progress is steadily going up. That means some people are getting tasks, just that if your computer is not asking for them at exactly the right time it doesn't get them. With a bit of luck when some of the new regional models come out of beta there will be enough to go round?

                            Post to thread

                            Message boards : Number crunching : "Project has no tasks available"




                            Copyright © 2002-2014 climateprediction.net