Announcement

Collapse
No announcement yet.

We're #200 @ Rosetta@Home, improving the world!

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

  • We're #200 @ Rosetta@Home, improving the world!

    More news to follow...
    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
    #198 now. We're doing fine with a #84 by RAC but we've had better times. Only 6 of the 13 that ever joined actually earn credits nowadays and if Nicram does not up his game, I'll be #1 again soon(tm).

    We'll need to about triple our credits to get to #100 so we can use all the help we can get. Had to build my oldest a new computer which is now adding credits as well. Had to take it off her previous machine as it was slowing it down, mostly at boot time as Rosetta reads a lot of small files. SSDs annihilate that issue though.

    Personally, I am now #24 of the Netherlands by TC and #4 by RC. If just my other two kids would apply themselves a bit more I'd have a great excuse to build them the same machine and get to #1 by RAC of the Netherlands...
    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


    • #3
      It looks like we will take over a couple of teams in the next few days, but then it will take a bit longer...

      MMM might help a bit when winter in Australia happens...

      I apparently lost a bit of RAC, but am not sure why: computer was on, and was not doing anything... Perhaps it is the new virusscanner we got installed... Current weather here really prevents me from running on the dual Xeon... not that it contributes that much, but with 25° indoors, I don't want to add more heat...
      pixar
      Dream as if you'll live forever. Live as if you'll die tomorrow. (James Dean)

      Comment


      • #4
        As a team, we're losing RAC. We used to be close to 20K, not at 15 and going for 14. We could really use some more contributors.
        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


        • #5
          Wow, Nicram became a triple millionaire! Zokes is close to becoming a double millionaire.

          We're #195 now and 96 by RAC, we're losing ground in that respect. Some of the heavy crunchers appear to be phasing out I'm afraid...
          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
            Had to apply restriction on my server, allowing it it to run nights only. During the cold seasons, no prob, but during the warm season, it gets way too hot in that room.
            Titanium is the new bling!
            (you heard from me first!)

            Comment


            • #7
              Understandable. High summer I may have to do somethng similar (but I don't think so, the Server is at a good isolated location and it's underclocked and volted as well, we'll see).

              Anyway, enjoy the summer!
              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
                Hurray! We have a new member: box!. Welcome to the team, great! Funny that it says you joined since November 21st, 2007! (http://boinc.bakerlab.org/rosetta/sh...?userid=221948)

                Meanwhile, this disaster happened:
                Capture.PNG
                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


                • #9
                  Currently working on getting Boinc to play nice on a Ramdisk, as there's no longer a HDD for this sytem and I'd rather not have it thrash the remaining SSD's inside it. Once I get it working OK, I'll bring it back online for a more permanent level of work when it's on. Right now, just testing to see how best to make it work on the Ramdisk without losing work when it shuts down.

                  J1NG

                  Comment


                  • #10
                    Interesting. With my ASRock boards I got this X-FastRam ramdisk software that neatly saved on shutting down and loaded on boot. I also had another piece of software once, don't remember the name, that was free but limited to 4GB. What SW do you use?
                    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


                    • #11
                      Using Dataram Ramdisk, paid the license fee for it. Running a 32GB Ramdisk for scratch disk, page file and fast temp file storage. It too has a save on shutdown feature, but 32GB takes a small while to save, so I'd rather just leave it as a blank start up every time it restarts or cold boots.

                      Currently, it looks like I can manually have Boinc request new work units, but it incredibly decides to download some 30 new work units needing some 200 minutes each to complete. Need to find a way to limit how many it takes each go. Or I suppose I manually abort the excess work units.

                      *Update* Haven't found a way to manually set the amount of new work units I get yet, but the Abort option seems OK to use. It doesn't look like there's any negatives so far in using it.

                      J1NG

                      Comment


                      • #12
                        Try something like this:
                        Boinc Manager (advanced view) -> Tools -> Computing Preferences -> Network usage -> Set Mimimum work buffer to 0.01 days and Max additional work buffer to 0.1 days or somesuch. What are the values now?
                        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


                        • #13
                          Was at 0.00 and 0.25. Currently set to your settings, will see how it goes.

                          J1NG

                          Comment


                          • #14
                            Ah, in that case perhaps it is even better to set both at 0.00 or 0.00 and 0.01. Essentially, you do not want any tasks that it can not work on right now.

                            Also, if at all possible, you might do two Ramdisks, one for BOINC, say 2GB to 4GB or so and the rest for the other stuff. It could then perhaps save the BOINC-ramdisk on shutdown. Reasoning is, I'm not sure whether the project would suffer from aborts and this way it can always resume work I would think.

                            I do wonder what kind of system you have where you can spare 32GB as a RamDisk. Why can;t it be used as system ram? If it can, would it not make more sense to use it as system memory and disable a page file? Just wondering.
                            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
                              Originally posted by Umfriend View Post
                              Ah, in that case perhaps it is even better to set both at 0.00 or 0.00 and 0.01. Essentially, you do not want any tasks that it can not work on right now.
                              Cool, that sounds exactly like what I needed, can manage them on the Ramdisk easily then. Didn't want to manually abort them every time I was running them.

                              Also, if at all possible, you might do two Ramdisks, one for BOINC, say 2GB to 4GB or so and the rest for the other stuff. It could then perhaps save the BOINC-ramdisk on shutdown. Reasoning is, I'm not sure whether the project would suffer from aborts and this way it can always resume work I would think.
                              Not sure if they said it would cause issues, I assumed it was the same as a passed deadline. But this option is definitely better. Not entirely certain Dataram's Ramdisk allows two instances, but the option you've provided me allows me to control it much more easily, so shouldn't need to get another Ramdisk or another instance running anyway at this point.

                              I do wonder what kind of system you have where you can spare 32GB as a RamDisk. Why can;t it be used as system ram? If it can, would it not make more sense to use it as system memory and disable a page file? Just wondering.
                              Heh, it was designed to do a whole bunch of stuff early on. This is the same machine as before (64GB RAM) but lost the HDD when it failed. So it was used to handle: Video editing (Premiere, After Effects), Image manipulation (Photoshop), gaming (and why not? ), secure data handling (when needed), Virtual Machines (when needed, usually in association with secure data handling), Television recording/playback (don't have to worry about any sort of read/write jumping as it's going so fast) and anything else that would kill the SSD unecessarily was put onto the Ramdisk.

                              I could have left it as 64GB system memory and turned off the pagefile, but other than some of the above reasons, there were unfortunately some snags as a few installations required the damn thing to be present before I could get them installed. So rather than turn it off, just lumped it over to the Ramdisk and basically didn't need to ever remember it ever existed. Much easier than turning it on and off, etc.

                              J1NG

                              Comment

                              Working...
                              X