| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Different WUs? [resolved] |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 9
|
|
| Author |
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges: 
|
Dear team,
----------------------------------------I just got a WU which has a different naming convention than usual. The name is 20091203144517_ faah9423_ ZINC12021275_ xJ1_ xtal_ 01_ 0-- It has a deadline already after 4 days. The estimated running time seems also higher than ususal. Can anyone explain what is different with the new WUs? Why such a short deadline? Thanks. [Edit 1 times, last edit by mreuter80 at Dec 4, 2009 6:12:28 PM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
It's an experiment 31 http://fightaidsathome.scripps.edu/status
----------------------------------------as you probably already know, but can't tell you what the codes mean, but surely mnemonics for compounds in the database processed. Experiment 31: 0% Completed Experiment 31 involves screening the Otava library of approximately 360,000 "building blocks" against the "exo site" on the side surfaces of HIV protease (i.e., the putative allosteric inhibitor sites). This experiment will dock these fragments against the exo sites of 7 different targets, which include 5 carefully-selected snapshots from previous Molecular Dynamics simulations of the V82F/I84V multi-drug-resistant mutant "super bug." The other targets correspond to the two new fragment-bound crystal structures of HIV protease that were produced by our collaborator, Prof. C. David Stout at TSRI. These two new structures from Prof. Stout were also targeted in Experiments 25-28 and 30. This experiment involves faah8972 - faah11491. 4 days deadline means it's either a repair job or a replacement for a result that was not returned within the original 10 day deadline. For that see the Result Status page result detail. edit: The average run time have been trending up slightly, but experience tells that is normal variability while working through sometimes harder to compute docking/compound combinations. We see this on all WCG sciences at times Average Hours per job 11.24 7,82230 11.25 7,98416 11.26 7,98650 11.27 7,87248 11.28 7,71605 11.29 7,88040 11.30 8,14232 12.01 8,03719 12.02 8,01481 12.03 8,13220
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Dec 4, 2009 10:09:57 AM] |
||
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges:
|
Hi Sekerob,
----------------------------------------Thanks for trying to answer my question, but in this doesn't help. I know that we are well into experiment 31, I got also other WU from batch 9357 and lower. But you can see from this WU it is already batch 9423. Quite a jump and normally not happening. Please look at the name of the WU. Normal WU start with faah... This one starts with a date as it seems. Here is the quorum view of the WU: 20091203144517_ faah9423_ ZINC12021275_ xJ1_ xtal_ 01_ 0-- - In Progress 12/4/09 06:17:58 12/8/09 06:17:58 0.00 0.0 / 0.0 I don't see any indication that this is a repair job. Repair job have usually a deadline of 2 or 2.5 days. This one is 4 days. Maybe knreed or another technical team member can help. One comment on the running time: It seems that the running time is not much higher and does not fall out of the usually upper bandwidth (usually 4.5 hrs, this will probably run 5 hours). However, BOINC did an initial estimation of more than 11 hours. I know, one cannot trust these estimation, but the estimation on this machine for FAAH is usually not that far off. thanks again. [Edit 1 times, last edit by mreuter80 at Dec 4, 2009 1:10:09 PM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Yes, I've noticed this date prefix. It's a test result to find out how long the batch will run on volunteer machines. Based on the run times, the techs will either shorten or increase the number of docking attempts in the upcoming tasks for this batch. It's a routine procedure applied for FAAH on all batches... few actually notice them.
----------------------------------------edit: That may be the 3rd reason this task had a short deadline :D edit2: I've updated the deadline FAQ too to include the batch test result bit. The 4 day (40% of original deadline), has been in place for longer when a result of a mishap on the original project tasks.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 3 times, last edit by Sekerob at Dec 4, 2009 1:49:07 PM] |
||
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges:
|
Thanks Sekerob,
That explains the WU. But why is this not handled through the beta test project. I did not register for beta testing and don't want to participate in it. Also there are so many beta testers that are waiting for more beta test WU - they would be very happy to help out to run these WUs. Please don't get me wrong. I understand why that is needed. I just think this is not a good practice. I trust the WCG team to run good WUs (WU that are relevant for the experiment) and don't do any testing on my machine - that's what I agreed to. Thanks. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
They don't consider it beta because it only checking the time estimates for new experiment, not changes to the science software itself. They have been doing this for a really long time and only occasion someone notices it but all they ever notice is that names/ dates are different, they do not crash/ error out, or even timeout. The results are used for real, not just tossed away. Blink and the WU will be finished, points granted and you will likely never see one of these again. :up:
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Confirmed (said with assumed authority ;>).
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Sek and Snow's answers are correct. To enhance, there are 3 random work units chosen in the new batches that get sent out to reliable hosts. Once these are returned, we are able to do a quick time analysis on them to get what the average time of the batch will be. If these work units in a batch run for say 1 hour, then the split was over zealous but it is mainly used to prevent long 20-30 hour work units. Because the experiments change rather quickly, and there are many of them. This is the best practice for making sure the new experiments don't skew the estimation of the splits.
Hope this helps on elaboration :) -Uplinger |
||
|
|
mreuter80
Advanced Cruncher Joined: Oct 2, 2006 Post Count: 83 Status: Offline Project Badges:
|
Thanks for the clarification.
----------------------------------------I still think testing should be done with users who signed up for testing (just because one does something for a long time doesn't mean it is right or best practice), but since the result is used for the experiment I'm kinda OK with it. Thanks again to everyone answering my question. I will stop whining now  [Edit 1 times, last edit by mreuter80 at Dec 4, 2009 8:45:36 PM] |
||
|
|
|