| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Have very slow WU # FAHV_x3VQ7_IN_LEDGFa_rig_0220728_0005 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 234
|
|
| Author |
|
|
cpuandgpu
Cruncher Joined: May 14, 2014 Post Count: 3 Status: Offline |
I have a stock i7 3770k and i7 4770 that are taking around 24 hours. I'm running at 90% windows 7 64 bit.
|
||
|
|
cpuandgpu
Cruncher Joined: May 14, 2014 Post Count: 3 Status: Offline |
I have a stock i7 3770k and i7 4770 that are taking around 24 hours. I'm running at 90% windows 7 64 bit.
|
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1326 Status: Offline Project Badges: 
|
These long-running units are amazing especially runtimes for the I 7 4470 K. The way I look at it is it is good for people who are badge hunting. All the best to the people that have these to crunch on
---------------------------------------- |
||
|
|
Eric_Kaiser
Veteran Cruncher Germany (Hessen) Joined: May 7, 2013 Post Count: 1047 Status: Offline Project Badges: 
|
Just speaking for my personal situation. For me it is perfect to receive these wu as the first recipient. They have quorum and replication one. That means my androids will process them well despite of the long runtime and they are valid on return.
---------------------------------------- |
||
|
|
Mamajuanauk
Master Cruncher United Kingdom Joined: Dec 15, 2012 Post Count: 1900 Status: Offline Project Badges:
|
I've had a few of these 'LEDG' wu's the longest of which are:
----------------------------------------FAHV_ x3VQ7_ IN_ LEDGFa_ rig_ 0220833_ 0026_ 0-- Node23 Valid 04/09/14 22:17:53 07/09/14 21:20:49 43.27 / 45.49 569.6 / 569.6
Mamajuanauk is the Name! Crunching is the Game!
----------------------------------------  [Edit 2 times, last edit by Mamajuanauk at Sep 8, 2014 6:12:35 PM] |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges: 
|
FAHV_ x3VQ7_ IN_ LEDGFa_ rig_ 0220832_ 0025_ 0-- FWPC4 Error 04/09/14 13:04:45 08/09/14 15:09:22 61.17 / 61.21 64.8 / 0.0
----------------------------------------great, I error out after 60 hours and only ask for 1/16th of my normal points. Why are we still getting super long WU? i thought it was just one bad batch that wasn't cut properly. I still see many many WU that are in the 50, 60, 70 hour range.  |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Offline Project Badges: 
|
FAHV_ x3VQ7_ IN_ LEDGFa_ rig_ 0220832_ 0025_ 0-- FWPC4 Error 04/09/14 13:04:45 08/09/14 15:09:22 61.17 / 61.21 64.8 / 0.0 great, I error out after 60 hours and only ask for 1/16th of my normal points. Why are we still getting super long WU? i thought it was just one bad batch that wasn't cut properly. I still see many many WU that are in the 50, 60, 70 hour range. I think they are cut properly, but they incorporate that "flexibility" component which evidently makes them much more compute intensive. Maybe there is a way for the techs and/or scientist to make them shorter, but maybe the science requires these to be this way. We will only know when either one of them decides to comment further. Too bad these monsters can not be directed to only the higher end machines. BTW, did you get an error code for this one ? Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges:
|
no, I have only got 1 error so far. which is kinda strange since I have many that are over 70 hrs long and even in the 80s. I don't mind long WU, the problem is that it totally messed up my scheduling. The system can't go humming along thinking the WUs are 4 hours long, then comes several WUs that are 70 hours long. WUs timeout, other switch to high priority. 2 or 3 times the regular time is one thing, but 10-20 times? Yeah that is just not good.
---------------------------------------- |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Offline Project Badges:
|
Yes, I agree, these units do wreak havoc with the scheduling. The time to completion listed is completely erroneous and bears no resemblance to reality. The actual scheduling is really all out of whack, but I have not run short of work either. I am lucky in that I only keep 1 day cache so I have not had any go into panic mode. I will have some probably go in excess of 100 hours on a q6600 running Vista, but so far so good. Whenever they finish I will post. Looks like a couple days yet though.
---------------------------------------- Edit: Well the result is in on one of the units: FAHV_ x3VQ8_ B_ IN_ LEDGFa_ rig_ 0221586_ 0063_ 0-- Error 66.58 / 66.58 64.8 / 0.0 Reason Given: Maximum CPU time exceeded The number of tasks listed was 140 and I succeeded in getting 114 done. I estimate it would have taken about 80 hours to complete the work unit. The time limit appears to be set at 240,000 seconds, at least for this unit. I would have needed a limit set to perhaps 290,000 seconds. Disappointing  Cheers
Sgt. Joe
----------------------------------------*Minnesota Crunchers* [Edit 2 times, last edit by Sgt.Joe at Sep 9, 2014 11:47:42 AM] |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
Came here to say they were running over a day and a half on a Xeon server and then I see run times up to 80 hours.
----------------------------------------...slinks back into the shadows. You're right though. Something that typically runs 45 minutes to 1.5 hours should never jump to 36+ hours. This is not "more complex molecules" this is a gross error in work unit creation.  Distributed computing volunteer since September 27, 2000 |
||
|
|
|