Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
![]() |
World Community Grid Forums
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 98
|
![]() |
Author |
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I am a little unclear as to exactly which part of FAAH2 is giving you the headache. FAAH2 is designed to send a trickle message, a partial update of results if you will, every 10,000 steps. I'll gather more data and hopefully describe this better. What I'm seeing right now is that the estimated time remaining is several hours greater than the time until the deadline. (Yesterday I opted out of running FAAH2, but hadn't aborted a couple of WUs because it looked as if they would manage to finish by the deadline. The stats however, are now suggesting that that is unlikely.) The WUs often finish and are accepted, not sure exactly when, but I *think* after the deadline. (Alternatively, I suppose that the estimated time remaining might suddenly lose the last 5 hours or so out of 20.) I suspect that if computing happens after the deadline, then it isn't counted, for whatever reason. I don't care about credit; what worries me is that the biological computations might be wrong, perhaps involving wasted cycles, or perhaps designed by someone who hasn't done a good job of testing. If they haven't fixed these obvious things that look bad to us volunteers, do they know what they are doing with the chemistry? P.S.: This is not due to confusion over times, I do realize that the Boinc manager displays local time, and the online Results Status uses UTC. Here are some data. I had two FAAH2 WU's running. Here are lines from three screenshots from the Results Status page online (pipes added for clarity). The strange thing is that one of the two WUs continued running after it was awarded Valid status; it requested credit for that extra work (1.75 hours, 14.01 - 12.26), but it was not granted. 310.8 credit was claimed, but only the earlier 272.0 was granted. 1.75 hours of computing was apparently wasted. As mentioned above, this could produce a lower average credit per CPU hour for this project than for others. What was the WU doing for those 1.75 hours? Did it successfully report the results? I doubt that this is due to the default CPU % discussed in the next post, because these two WU's were running on the same machine (dual processors). I thought this behaviour might be because there's a time delay between credit being claimed and credit being granted, but now, nearly 12 hours later, the status of the strange WU still shows the same disparity, 310.8 claimed, 272.0 granted. Screen Shot 2015-12-19 at 17.45.21 Result name | Status | Time due/Return Time | CPU Time/Elapsed | Claimed/Granted BOINC Credit FAH2_avx38747-is_000065_0016_024_0-- | In Progress | 12/20/15 05:46:08 | 12.26 / 0.00 | 272.0 / 0.0 FAH2_avx38747-is_000021_0015_029_0-- | In Progress | 12/20/15 05:44:00 | 14.05 / 0.00 | 310.8 / 0.0 Screen Shot 2015-12-19 at 19.55.21 Result name | Status | Time due/Return Time | CPU Time/Elapsed | Claimed/Granted BOINC Credit FAH2_avx38747-is_000065_0016_024_0-- | Valid | 12/19/15 23:46:52 | 12.26 / 0.00 | 272.0 / 272.0 FAH2_avx38747-is_000021_0015_029_0-- | Valid | 12/19/15 23:46:45 | 15.85 / 0.00 | 349.6 / 349.6 Screen Shot 2015-12-19 at 21.52.05 Result name | Status | Time due/Return Time | CPU Time/Elapsed | Claimed/Granted BOINC Credit FAH2_avx38747-is_000065_0016_024_0-- | Valid | 12/19/15 23:46:52 | 14.01 / 0.00 | 310.8 / 272.0 FAH2_avx38747-is_000021_0015_029_0-- | Valid | 12/19/15 23:46:45 | 15.85 / 0.00 | 349.6 / 349.6 [Edit 3 times, last edit by Former Member at Dec 20, 2015 3:24:14 PM] |
||
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
One of the 'gotcha' is that the WCG client comes default with a 50 or 60% CPU time setting. If a tasks then at start is projected to need e.g. 60 hours, but the used computing time is actually 100/60th of 100/50th, then the 'clock' time will be going past the deadline. Siuch runtimes are way outside the project average of below 16 hours... see chart http://bit.ly/WCGART1 where current average is projected to be below 13 hours.
|
||
|
Wee Todd Didd
Cruncher Joined: Jan 7, 2009 Post Count: 3 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Sorry FightAIDS@Home Phase 2 project, I have to bail on this one.
I find it near impossible to complete this project in the time given. My PC's do not run 24/7. If the time to return results ever gets extended I will return. |
||
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
Were it not that every extension will cause the end result the scientists need after 30 or 40 iterations is moving out that much further, so don't feel sorry if you have to bail on volunteering for this research work.
|
||
|
esteban521
Cruncher Joined: May 22, 2008 Post Count: 4 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() |
This must be the wrong thread but I don´t know where else to ask, sorry
I had a laptop with an intel core2duo p8800, now I have a core i7 2670QM, it IS somewhat dated now but the core2duo was faster than the i7 and I don´t know why , does anyone know? is it something I have to configure? Thnak you |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Yes, is a bit off topic, maybe start a new thread in "BOINC Agent Support".
Taking a stab in the dark, it maybe something to do with the TurboBoost not kicking in on your i7 when WCG is running. Get a new thread started, so we can keep this one on-topic and I'm sure lots of helpful suggestions will pour in :) |
||
|
esteban521
Cruncher Joined: May 22, 2008 Post Count: 4 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() |
Will do, thank you
|
||
|
wolfman1360
Senior Cruncher Canada Joined: Jan 17, 2016 Post Count: 176 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Hi,
----------------------------------------Jumping in a bit belatedly on this one, however I'm curious. What, exactly, is the reason for cashing days of work? What does it accomplish? Or is it simply so more offline work can be done without internet access? Very, very curious. Even on my lowest end atom machine (d2550), I have yet to see anything about high priority on any project at the default 0.5 days.
Crunching for the betterment of human kind and the canines who will always be our best friends.
AWOU! |
||
|
|
![]() |