| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Question about FAHV tasks and FreeBSD OS |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 9
|
|
| Author |
|
|
boinc127
Cruncher Joined: Jan 28, 2012 Post Count: 11 Status: Offline Project Badges: 
|
I am currently running WCG FAAH and FAHV tasks using FreeBSD with the Linuxulator. All of my FAAH tasks are being completed and are all valid, however for some reason all of the FAHV tasks are being completed, but all of the tasks are marked invalid. The only solution I have used so far is to abort the FAHV tasks and keep the FAAH tasks, however this doesn't seem like it would work out the best on WCG's end. Is there a way I can specify only the FAAH tasks in an app_info.xml file or something similar?
----------------------------------------[Edit 2 times, last edit by boinc127 at Mar 10, 2014 2:53:33 PM] |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7852 Status: Offline Project Badges: 
|
Could you post the result log for one of the invalid FAHV units ? Also there is no way of which I am aware to specify either Vina or Autodock units.
----------------------------------------Thanks
Sgt. Joe
----------------------------------------*Minnesota Crunchers* [Edit 1 times, last edit by Sgt.Joe at Mar 11, 2014 12:10:52 AM] |
||
|
|
boinc127
Cruncher Joined: Jan 28, 2012 Post Count: 11 Status: Offline Project Badges:
|
No problem at all... and thank you for the prompt reply.
----------------------------------------https://secure.worldcommunitygrid.org/ms/devi....do?workunitId=1010071884 Project Name: FightAIDS@Home - Vina Created: 02/28/2014 01:24:53 Name: FAHV_x3VQE_A_IN_FBPb_rig_0200695_2294 Minimum Quorum: 2 Replication: 3 Result Log Result Name: FAHV_ x3VQE_ A_ IN_ FBPb_ rig_ 0200695_ 2294_ 1-- <core_client_version>7.2.33</core_client_version> <![CDATA[ <stderr_txt> INFO: No state to restore. Start from the beginning. [14:36:54] Number of tasks = 7 [14:36:54] Starting task 0,CPU time is 0.000000. [14:36:54] ./ZINC16932416.pdbqt size = 20 4 ../../projects/www.worldcommunitygrid.org/fahv.x3VQE_A_IN_FBPb_rig.pdbqt size = 2521 0 [14:43:55] Finished task #0 cpu time used 1.024184 [14:43:55] Starting task 1,CPU time is 1.024184. [14:43:55] ./ZINC16932420.pdbqt size = 20 4 ../../projects/www.worldcommunitygrid.org/fahv.x3VQE_A_IN_FBPb_rig.pdbqt size = 2521 0 [14:50:50] Finished task #1 cpu time used 1.037254 [14:50:50] Starting task 2,CPU time is 2.061438. [14:50:50] ./ZINC16932433.pdbqt size = 21 1 ../../projects/www.worldcommunitygrid.org/fahv.x3VQE_A_IN_FBPb_rig.pdbqt size = 2521 0 [14:56:44] Finished task #2 cpu time used 0.895479 [14:56:44] Starting task 3,CPU time is 2.956917. [14:56:44] ./ZINC16932442.pdbqt size = 21 6 ../../projects/www.worldcommunitygrid.org/fahv.x3VQE_A_IN_FBPb_rig.pdbqt size = 2521 0 [15:06:02] Finished task #3 cpu time used 1.107562 [15:06:02] Starting task 4,CPU time is 4.064479. [15:06:02] ./ZINC16932446.pdbqt size = 24 5 ../../projects/www.worldcommunitygrid.org/fahv.x3VQE_A_IN_FBPb_rig.pdbqt size = 2521 0 [15:15:41] Finished task #4 cpu time used 1.125573 [15:15:41] Starting task 5,CPU time is 5.190052. [15:15:41] ./ZINC16932450.pdbqt size = 24 5 ../../projects/www.worldcommunitygrid.org/fahv.x3VQE_A_IN_FBPb_rig.pdbqt size = 2521 0 [15:25:03] Finished task #5 cpu time used 1.119533 [15:25:03] Starting task 6,CPU time is 6.309585. [15:25:03] ./ZINC16932453.pdbqt size = 29 6 ../../projects/www.worldcommunitygrid.org/fahv.x3VQE_A_IN_FBPb_rig.pdbqt size = 2521 0 [15:37:57] Finished task #6 cpu time used 1.224218 15:37:57 (55822): called boinc_finish </stderr_txt> ]]> Not only is the result invalid, it causes my wingman to have to wait even longer to validate their results because the server has to reassign another computer to crunch the correct solution to the workunit. I'm guessing its messy for the back end of WCG and probably screwing up statistics, etc. But I would love to keep crunching these workunits while I'm using FreeBSD (I switch between FreeBSD and Windows on the same computer, depending on the work I'm doing). 100% of the workunits from FAHV are invalid as crunched on my computer. Now that I look at the run time compared to other computers, the workunits run about an hour less than the other computers, but don't error out. [Edit 1 times, last edit by boinc127 at Mar 11, 2014 12:46:23 AM] |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7852 Status: Offline Project Badges:
|
It looks like your FAHV units are ending after only a few seconds on each task in the workunit. This is definitely not normal. There is no indication in the result file as to why this is happening. Perhaps another member running FreeBSD could offer some guidance. I would not lose any sleep over the items needing to be re-issued, because the system is built to handle this contingency. However, if they are all erroring, I can see why this poses a problem and creates extra work. Hopefully someone else can help with your problem on these.
----------------------------------------Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
boinc127
Cruncher Joined: Jan 28, 2012 Post Count: 11 Status: Offline Project Badges:
|
Thanks very much for the assistance. I will get around to trying to debug this or figuring out exactly whats wrong sometime. FreeBSD can be quite quirky sometimes, especially running newer linux programs. I'm not going to worry too much about it, I just hate to muck up the WCG backend servers with a bunch of aborted units. I can always run the Energy workunits, they seem to give me no problems at all.
Cheers, Jeremy |
||
|
|
branjo
Master Cruncher Slovakia Joined: Jun 29, 2012 Post Count: 1892 Status: Offline Project Badges: 
|
Vina is strange. I am receiving lot of invalids on my Mac OS X (also already finished projects like GFAM, SN2S and DSFL), crunchers with AMD CPU's also reporting lot of invalids.. So I am trying to avoid Vina applications crunching on my Mac
---------------------------------------- Cheers   Crunching@Home since January 13 2000. Shrubbing@Home since January 5 2006  |
||
|
|
armstrdj
Former World Community Grid Tech Joined: Oct 21, 2004 Post Count: 695 Status: Offline Project Badges:
|
There is no way to only get work for one of the FAAH applications, however over time you should start getting more FAAH AutoDock work over FAAH VINA because you are returning invalids for VINA. VINA is different than AutoDock in that it uses the compound application model. There is a main application (wrapper) which controls the job flow and forks multiple instances of the VINA applicaiton as directed by the workunit. For you it looks like the wrapper is running as expected but for some reason the VINA applicaiton is running for an unexpected very short duration but not producing an error. I am not sure why this is the case.
Any other FreeBSD users have any ideas? Thanks, armstrdj |
||
|
|
boinc127
Cruncher Joined: Jan 28, 2012 Post Count: 11 Status: Offline Project Badges:
|
How bizarre, I saw your information concerning the adaptive ability of the WCG servers (about returning invalid VINA workunits and only returning valid AutoDock units) would cause the server to send more AutoDock units instead of VINA... so I decided to run a few more VINA workunits instead of aborting all of them. Anyway, one of my VINA workunits was actually validated. Now I'm confused. I hope I'm not returning invalid results, and my wingman returned invalid results as well, and it was counted as valid.
Project Name: FightAIDS@Home - Vina Created: 03/15/2014 20:10:26 Name: FAHV_x3NF6_A_IN_LEDGFa_rig_0204722_2587 Minimum Quorum: 2 Replication: 3 Result Name App Version Number Status Sent Time Time Due / Return Time CPU Time / Elapsed Time (hours) Claimed/ Granted BOINC Credit FAHV_ x3NF6_ A_ IN_ LEDGFa_ rig_ 0204722_ 2587_ 3-- 706 Valid 3/17/14 06:59:15 3/17/14 22:16:20 0.00 1.1 / 29.2 FAHV_ x3NF6_ A_ IN_ LEDGFa_ rig_ 0204722_ 2587_ 2-- 706 Invalid 3/16/14 21:25:24 3/17/14 06:58:40 2.72 82.2 / 14.6 FAHV_ x3NF6_ A_ IN_ LEDGFa_ rig_ 0204722_ 2587_ 1-- - Detached 3/16/14 19:04:41 3/16/14 21:16:53 0.00 0.0 / 0.0 FAHV_ x3NF6_ A_ IN_ LEDGFa_ rig_ 0204722_ 2587_ 0-- 706 Valid 3/16/14 19:03:35 3/16/14 23:13:12 0.00 57.4 / 29.2 Result Name: FAHV_ x3NF6_ A_ IN_ LEDGFa_ rig_ 0204722_ 2587_ 0-- <core_client_version>7.2.33</core_client_version> <![CDATA[ <stderr_txt> INFO: No state to restore. Start from the beginning. [16:23:40] Number of tasks = 3 [16:23:40] Starting task 0,CPU time is 0.000000. [16:23:40] ./ZINC30145448.pdbqt size = 29 6 ../../projects/www.worldcommunitygrid.org/fahv.x3NF6_A_IN_LEDGFa_rig.pdbqt size = 2830 0 [17:02:23] Finished task #0 cpu time used 1.161138 [17:02:23] Starting task 1,CPU time is 1.161138. [17:02:23] ./ZINC30164496.pdbqt size = 22 5 ../../projects/www.worldcommunitygrid.org/fahv.x3NF6_A_IN_LEDGFa_rig.pdbqt size = 2830 0 [17:28:23] Finished task #1 cpu time used 1.134199 [17:28:23] Starting task 2,CPU time is 2.295337. [17:28:23] ./ZINC30171260.pdbqt size = 29 8 ../../projects/www.worldcommunitygrid.org/fahv.x3NF6_A_IN_LEDGFa_rig.pdbqt size = 2830 0 [18:13:02] Finished task #2 cpu time used 1.198277 18:13:02 (2751): called boinc_finish </stderr_txt> ]]> Perhaps the workunit validated because the number of tasks assigned to the workunit were so small? |
||
|
|
armstrdj
Former World Community Grid Tech Joined: Oct 21, 2004 Post Count: 695 Status: Offline Project Badges:
|
I would not worry about your results, if they are marked valid then they have been verified on our end and are good. I looked back at your log and one issue appears to be that your returned cpu time is not correct. I thought they were running extremely short but looking at the time stamps in the log the elapsed time is not short. The incorrect cpu time should not be related to valid/invalid results. There are two different methods used to calculate cpu time, one is to use a linux system call getrusage and the other is to use the /proc filesystem. I believe the current version of the VINA application uses the /proc file system. I would continue to run Fight Aids @ Home if you would like and keep an eye on results. Over time if you continue to produce more invalids on the VINA application you should start getting more AutoDock workunits.
Any other FreeBSD users have issues with returning the incorrect CPU time for any projects and have any advice? Thanks, armstrdj |
||
|
|
|