| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: CPU check point |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 4
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello,
I'm using BOINC software on my Mandriva Linux, and my projects are Einstein@home and World Community Grid (FAAH only). I'm using command line version - not Boinc manager. When I check my results status (./boinc_cmd --get_result) the CPU check point on WCG result shows very little time: - CPU check point time 63.619328 and current CPU time 3570.867145 or - CPU check point time 7607 and current CPU time 10762, When I swith off boinc (./boinc_cmd --quit) and swith it on again it shows very little current CPU time - the same as check point CPU time. Some time of my work is lost  By the way - Einstein@home's checkpoints are very close to current CPu time. My questions: Why checkpoints are made so rare ? Why there are no information about servers status - (I woluld like to know something about fraction of FAAh done) Greetings from Polnad  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello Kwiatek,
Why checkpoints are made so rare ? The checkpoints are initiated by the application software which is provided to the World Community Grid by the scientists running the project (Rosetta for the HPF project or Autodock for the FAAH project). Rosetta appears to take frequent checkpoints and you will not lose much when you need to shutdown and restart. This is not the case for Autodock as you have seen. The developers at the World Community Grid are aware of this and have indicated that a change is planned to cause the application to checkpoint more frequently, but I am not sure when this change might be made. Why there are no information about servers status - (I woluld like to know something about fraction of FAAh done) Usually the best information regarding how far along a project is can be found in the forum for that project. For FAAH, that forum can be found here. The scientists that are behind the FAAH project have not posted that frequently in our forums, but they do have a web site for the project which may also provide you with some of the information you are looking for here |
||
|
|
retsof
Former Community Advisor USA Joined: Jul 31, 2005 Post Count: 6824 Status: Offline Project Badges: 
|
As I understand it, when running BOINC it is not yet possible to pick the particular World Community Grid project to run. It is possible that you will get a HPF/Rosetta at random instead of FightAids@Home for the next workunit. This might be next time, or later. There is some talk about making the selectable, but that hasn't been figured out yet.
----------------------------------------Yes, Autodock checkpoints seem to be more variable than HPF/Rosetta. Even those have some variance. I have seen some on both projects as quick as every 2-3 minutes apart on a fast computer. I had a couple of FA@H workunits that took only one intermediate checkpoint --- at 50% completion. An hour here could be 2-3 hours on a slow machine. This has come into play when part-timers turn off their computers ... and the workunit restarted at the beginning when they resumed. All we can recommend is to run it for a greater percentage of the day, or 24/7 ideally. I would also suspect that this could be a problem if any one project has too small of a BOINC percentage. It doesn't run enough to get to a checkpoint before going on to the next project. That is a general comment, not for WCG in particular.
SUPPORT ADVISOR
----------------------------------------Work+GPU i7 8700 12threads School i7 4770 8threads Default+GPU Ryzen 7 3700X 16threads Ryzen 7 3800X 16 threads Ryzen 9 3900X 24threads Home i7 3540M 4threads50% [Edit 4 times, last edit by retsof at Mar 14, 2006 2:37:33 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Rick Alther, the programmer who boards the project applications to run on the WCG, posted a good explanation about FAAH check points here: http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=6294#51277
The problem is that AutoDock offers very few points where it is easy to save the state of the program other than by saving everything. [266 MB, 300 MB ? Aargh!!] My understanding of project selection is - - uncertain. But that never stops me, so here goes!  When we activate a second project for a platform on the server, we switch to a multi-project section of code (an add-on kludge) and the current project selections get reset to 'all projects'. Most recently, this happened to BOINC when we added FAAH to it. You can go to 'My Grid' - 'My Projects' and once again select a specific project (FAAH or HPF) if you want to do so.There is a permanent difference between the UD and BOINC platforms in that it is also possible to select specific projects for individual computers (devices) using Device Manager for the UD platform. BOINC is all or nothing. You make the same project choice for all of your BOINC computers. Except that you can have different profiles and can choose to run WCG or not on each computer, but you cannot select a specific WCG project for a specific BOINC computer. Still, that leaves a lot of room for choice (or confusion). Lawrence |
||
|
|
|