Announcement

Collapse
No announcement yet.

We're #150 @ Rosetta@Home, improving the world

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • We're #150 @ Rosetta@Home, improving the world

    He VJ, you've been out for over a week now, is all well?
    Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
    [...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen

  • #2
    Hmm.. Client is up, and showing "waiting to run"... Again the issue from last time that surfaced. I don't need much computing power now, so I'll restart the virtual machine and set Boinc to "calculate always", to see if it will catch on again...

    edit: looks like it is not starting... just waiting...
    Last edited by VJ; 22 April 2015, 05:49.
    pixar
    Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

    Comment


    • #3
      That is a linux vm, no? I know nothing about those kind of setups

      Normally, waiting to run implies that a resource is in demand beyond the threshold, usually the cpu. Is there something like task manager that you can run within the vm to check what the BOINC client is seeing that would cause it to halt the workunits?

      BTW, is may be that Mashmallowman is back up running, that;d be great!
      Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
      [...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen

      Comment


      • #4
        Originally posted by Umfriend View Post
        Normally, waiting to run implies that a resource is in demand beyond the threshold, usually the cpu. Is there something like task manager that you can run within the vm to check what the BOINC client is seeing that would cause it to halt the workunits?
        When it is CPU, the message is "Computation suspended - CPU in use"...
        There is nothing else running on the VM, and it is assigned 2 CPUs... I cannot find anything that may cause the BOINC client to suspend calculation.

        The VM was assigned 512 MB (which is the default for Debian), I've now increased it to 1024 MB, perhaps the resource you refer to is memory...?
        pixar
        Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

        Comment


        • #5
          Ah yes, of course, my bad. I play around with max CPUs on my lappy and it will say waiting once I squeeze the number of CPU's.... And yes, AFAICS, WUs tend to take anywhere between 300 and 510 MB and I assume that part of that 512MB will be used for other stuff so this might well make a difference.
          Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
          [...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen

          Comment


          • #6
            Increasing the memory seems to have done the trick: it started computing. I've changed the setting back to only work at night (I need computing time), but it should be adding things soon™.

            Now if there could be a way to set it to crunch on 1 CPU during the day and 2 CPUs at night....
            pixar
            Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

            Comment


            • #7
              Originally posted by VJ View Post
              Now if there could be a way to set it to crunch on 1 CPU during the day and 2 CPUs at night....
              Amen! Or, for instance, that with an increasing CPU load, it would suspend WUs one at a time instead of the all-or-nothing approach. Ah well, with you back on trakc and MMM going on winter we're good.

              @Nicram, big words but for now I am increasing my lead
              Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
              [...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen

              Comment


              • #8
                Originally posted by Umfriend View Post
                Amen! Or, for instance, that with an increasing CPU load, it would suspend WUs one at a time instead of the all-or-nothing approach. Ah well, with you back on trakc and MMM going on winter we're good.
                I might setup a second VM with a single CPU, and schedule that one to run during the day. But it will have to be later, I have some deadlines to reach. The current VM seems to be working now, so it must have been a memory issue.
                pixar
                Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

                Comment


                • #9
                  One Q though, why run it in a VM? I would assume that running it on the "main" machine would cause the WUs to suspend and offload when you need the resources?
                  Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
                  [...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen

                  Comment


                  • #10
                    For some reason, running it normally slows down the system. Suspending and offloading the resources goes much too slow (literally MINUTES that the system is completely unresponsive, and even simple operations get delayed a lot). I suspect it is BOINC+PC related, because I don't have such a big issue with other programs. As VMWare behaves nicer when it comes to offloading and sharing resources, I decided to go that route. It costs me a bit of flexibility (e.g. no automatic offloading), but at least my system behaves nicer.
                    pixar
                    Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

                    Comment


                    • #11
                      Could it be there is not enough memory allocated for two large WU's? You may need like 1.2GB or so? Can't see when the VM last contacted the Rosetta server.
                      Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
                      [...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen

                      Comment


                      • #12
                        It was not switched on this weekend: I need to run some important simulations, that also ran overnight.
                        pixar
                        Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

                        Comment


                        • #13
                          I think it reported... Finally...
                          pixar
                          Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

                          Comment


                          • #14
                            It has and we're finally at #149. Progress is slow nowadays. To bad we can;t get some of the others to crunch as well....OH WAIT, RC Agent just joined! Welcome and many thx! That makes 7 actively crunching!
                            Join MURCs Distributed Computing effort for Rosetta@Home and help fight Alzheimers, Cancer, Mad Cow disease and rising oil prices.
                            [...]the pervading principle and abiding test of good breeding is the requirement of a substantial and patent waste of time. - Veblen

                            Comment


                            • #15
                              Breaking the 100 barrier looks like it'll take a while. Probably looking at years ahead unless some of them drop out entirely.

                              Would have had a better earlier run on my end if the multipliers on the main CPU worked as they should. The amount of crunching I did prior to a break back a year ago, should have been around 300k credit; instead only netted around 150k and only just gone by 300k recently (375k of this post).

                              Only consolation is that I'm now doing some 100k a month, so should be breaking the 1m barrier by Christmas so long as I keep the current rate going.

                              J1NG

                              Comment

                              Working...
                              X