| 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: 67
|
|
| Author |
|
|
wplachy
Senior Cruncher Joined: Sep 4, 2007 Post Count: 423 Status: Offline |
For the users seeing the very low efficiency is this only with running multiple workunits at the same time? Is anyone seeing this low efficiency while only running one CEP2 workunit? Thanks, armstrdj Yes, I'm seeing it on two (2) different 2-core devices that are running only a single WU each. BETA_E200532_749_A.25.C19H11N3S2Se.175.0.set1d06_1 Elapsed=12:24:12 CPU=02:37:08 BETA_E200531_796_A.26.C20H11N3S2Si.82.1.set1d06_0 Elapsed=20:27:33 CPU=07:17:17 Both devices are and have been showing CPU usage between 92%-100% for the WUs. Bill P
Bill P
![]() |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
For the users seeing the very low efficiency is this only with running multiple workunits at the same time? Is anyone seeing this low efficiency while only running one CEP2 workunit? Thanks, armstrdj Aboslutly, I have a 2.4 single core no HT P4 and boinctasks is saying it is running at 10% cpu efficiency. Here is the unit: BETA_E200532_795_A.24.C19H13NS2SeSi.180.3.set1d06_1 using beta11 version 637 Been doing this a lot ever since this project started. |
||
|
|
codes
Advanced Cruncher Joined: Oct 20, 2009 Post Count: 142 Status: Offline |
Intel Mac, Intel Core 2 Duo CPU T7700 @ 2.40GHz.
BETA_ E200532_ 036_ A.27.C19H9N3O3S2.7.0.set1d06_ 0-- 637 Pending Validation 11/17/10 12:01:02 11/17/10 19:52:46 3.46 Application exited with RC = 0x100 [13:24:32] Finished Job #12 [13:24:32] Starting job 13,CPU time has been restored to 12337.494677. [13:24:32] Skipping Job #13 [13:24:32] Starting job 14,CPU time has been restored to 12337.494677. [13:24:32] Skipping Job #14 [13:24:32] Starting job 15,CPU time has been restored to 12337.494677. [13:24:32] Skipping Job #15 called boinc_finish This one ran and exited in about half the time the production WUs do. It was crunching concurrently with a HCMD2 WU. |
||
|
|
yose-ue
Cruncher Joined: Dec 27, 2008 Post Count: 21 Status: Offline Project Badges:
|
For the users seeing the very low efficiency is this only with running multiple workunits at the same time? Is anyone seeing this low efficiency while only running one CEP2 workunit? Extreme loss of time occurs regardless of how many cep2 units you process at a time. For those of us that have the problem it happens every time. The time loss occurs at or around the time of a checkpoint. It is most noticeable at the end of 2 because that is the largest job. When job 2 finishes the percent complete will drop from 30 to 40 percent down to maybe 5 percent and the cpu time will drop from 3 to 4 hours down to less than a half hour. I have leave in memory checked but that is not an issue since it is not starting over. The workunit continues to crunch with no interruption. The version of linux core will correct the problem If I use 2.6.32-24 or earlier I do not have any problems crunching cep2 workunits. I am certain that it is more than just using a more recent kernel since most people seem to be able to crunch cep2 using more recent versions of linux kernel. Hardware also seems varied between those of us that have had trouble with these huge loss of credit for the time spent. One person was even trying different linux distributions with no luck until he went with an older kernel. There is more discusion in the "poor crunching on this project" thread under cep2. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
For the users seeing the very low efficiency is this only with running multiple workunits at the same time? Is anyone seeing this low efficiency while only running one CEP2 workunit? The report I initially filed noted all 3 machines were running just one WU apiece, and after that report I changed the 'restrictions' setting to allow one per core. With one per core this CEP2 beta came close to overheating the Q6600 (GKrellm reported its CPU temp at 158 and started beeping an alarm at me; with one WU per core running on the Einstein@home GC-S5R4 application, it's currently right at 138F). The temps of the 2 AMD machines did not elevate nearly that much... about 5 degrees on the dual core, and 10 degrees on the Phenom II... over what they usually run with a mix of Einstein, Orbit, Docking, HPF2 and FAAH. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I also do not see any help for the extreme time loss -- if anything it might be worse. BETA_ E200531_ 212_ A.26.C20H11NO2S3.217.4.set1d06_ 1-- shows 2.84 hours for me, 9.52 for my wingman. This was run on my slow Intel box, it was somewhere over 11 hours elapsed time.
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
The final report, Linux Ubuntu 10.04.01 with kernel 2.6.32.26 bld 25, 11 results, gap times suffixed. Notably, those with the shortest gaps/highest efficiency, ran after the GUI crashed and did a Alt-F1 to get into a terminal and did some bashing to get the wifi going again and then monitored from remote. It's amazing how much a GUI can eat in time, so after digging "init 3" will be the command to run when not needing the glossing front end:
----------------------------------------6.37 beta11 BETA_E200532_524_A.27.C18H9N7S2.5.0.set1d06_1 08:07:35 (07:54:19) 18-11-2010 09:29 18-11-2010 09:35 Reported: OK (u) 13 min. 6.37 beta11 BETA_E200532_401_A.26.C20H11NO2S3.174.3.set1d06_2 07:54:44 (07:40:04) 18-11-2010 08:31 18-11-2010 08:37 Reported: OK (u) 14 min. 6.37 beta11 BETA_E200532_266_A.26.C20H11NO2S3.31.3.set1d06_1 08:19:36 (07:49:48) 18-11-2010 01:06 18-11-2010 01:14 Reported: OK (u) 30 min. 6.37 beta11 BETA_E200532_236_A.24.C19H13NS2SeSi.118.1.set1d06_0 10:00:54 (09:11:17) 17-11-2010 23:20 17-11-2010 23:29 Reported: OK (u) 49 min. 6.37 beta11 BETA_E200531_409_A.26.C20H11NO2S3.215.2.set1d06_0 08:38:04 (07:58:31) 17-11-2010 16:40 17-11-2010 16:47 Reported: OK (u) 40 min. 6.37 beta11 BETA_E200531_890_A.24.C19H13NS2SeSi.209.0.set1d06_1 07:26:12 (07:04:07) 16-11-2010 22:51 16-11-2010 23:51 Reported: OK (u) 22 min. 6.37 beta11 BETA_E200530_286_A.24.C23H18S.13.set1d06_0 04:23:34 (04:03:38) 16-11-2010 19:33 16-11-2010 19:39 Reported: OK (u) 20 min. 6.37 beta11 BETA_E200530_256_A.24.C18H10N2S2Se2.24.1.set1d06_0 06:56:54 (06:40:12) 16-11-2010 11:59 16-11-2010 13:53 Reported: OK (u) 16 min. 6.37 beta11 BETA_E200530_249_A.24.C19H12S3SeSi.11.1.set1d06_1 08:46:22 (08:24:19) 16-11-2010 11:59 16-11-2010 16:15 Reported: OK (u) 22 min. 6.37 beta11 BETA_E200530_147_A.24.C23H18S.10.3.set1d06_1 07:44:13 (07:18:02) 16-11-2010 00:43 16-11-2010 07:27 Reported: OK (u) 26 min. I'll be waiting on the 2.6.38 miracle kernel with new scheduler as "using" Linux with GNOME or Unity of KDE and all these front ends do not seem to bring out the best of the platform... good for industrial headless crunching. FAIC, go with this release ASAP, regardless if not able to fix these weird excessive timelosses that some see with (un) certain kernel/hardware combos. 93% efficiency or better than 95% efficiency. AND this was running always concurrent with a 2 Beta or CEP2 / 2 Other combination on the quad for the whole of the Beta test.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
yose-ue
Cruncher Joined: Dec 27, 2008 Post Count: 21 Status: Offline Project Badges:
|
I have one of the systems that have the extreme time loss and I don't think that this problem should prevent this bata version from becoming the release version. The extreme efficiency problem is the same between the current release and the bata.
I still hope you will continue to work on this problem for future releases. This is just a shot in the dark. My installation is just a stock installation except I get my INTERNET from a city wide wireless service. There is not much support for wireless devices in linux. I needed to us my windows driver using Ndiswrapper which allows you to use windows wireless drivers in linux. Just a hope that any information that may help locate the problem will help solve it. |
||
|
|
X-Files 27
Senior Cruncher Canada Joined: May 21, 2007 Post Count: 391 Status: Offline Project Badges:
|
My results (all cores/threads running CEP2 beta):
----------------------------------------6.37 Beta - The Clean Energy Project - Phase 2 BETA_E200531_805_A.26.C17H9N7SSe.2.0.set1d06_1 07:17:35 (07:08:41) 11/16/2010 07:49:15 PM 11/16/2010 07:56:01 PM Reported: OK (u) 9minGap times is now comparable with windows - seeing single digit. Results is running under Ubuntu 10.10 server with 2.6.36 kernel ![]() ![]() |
||
|
|
mikaok
Senior Cruncher Finland Joined: Aug 8, 2006 Post Count: 489 Status: Offline Project Badges:
|
6.37 Beta - The Clean Energy Project - Phase 2 BETA_E200531_805_A.26.C17H9N7SSe.2.0.set1d06_1 07:17:35 (07:08:41) 11/16/2010 07:49:15 PM 11/16/2010 07:56:01 PM Reported: OK (u) 9min
to infinity and beyond
![]() |
||
|
|
|