Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
World Community Grid Forums
Category: Completed Research Forum: The Clean Energy Project - Phase 2 Forum Thread: CEP2 work units to have 18 hour limit |
No member browsing this thread |
Thread Status: Active Thread Type: Sticky Thread Total posts in this thread: 48
|
Author |
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
The TTC [Time to complete], are consistently and utterly incorrectly represented in the client. Some merciful day the technicians fix this by telling the client the cap is the cap is the cap: 18 hours, no matter the speed of the device. In a long 'agone day I participated at Rosetta. There you could tell how long you wanted to run your tasks, and that is what you got, cut to size. If you asked for 6 hours of folding, you got 6 hours of folding, not 4 or 2 or 12, just 6 exactly 6, in short it seems possible. Blame part of it on the abused <dont_use_dcf> DCF=1.000000 lock. Clients don't learn anymore, they get the production average runtime pushed through, where the DCF 1.0 is a global WCG value, not a project level.
|
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have reorganized my projects. Now, only my fastest processor will get CEP2 WUs. I hope to add one more computer soon and based on it's performance I hope to be able to add CEP2 to it.
|
||
|
kokitch
Cruncher Joined: Jun 17, 2009 Post Count: 1 Status: Offline Project Badges: |
And what exactly is the point of this time limit? I never had errors with CEP before and in the past week I have 10 tasks that have errored out because of the limit. Imagine all the wasted electricity!
|
||
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7219 Status: Offline Project Badges: |
And what exactly is the point of this time limit? I never had errors with CEP before and in the past week I have 10 tasks that have errored out because of the limit. Imagine all the wasted electricity! Well, one point of the time limit is to prevent a WU with an infinite loop from running forever. Another reason is to prevent a WU which may be diverging rather than converging to have a definite end. A third reason might be if the molecule is too big, it won't hog your machine for several days while it completes. I am sure there are other reasons also. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
Procyon Lotor One
Cruncher Joined: Jun 28, 2008 Post Count: 14 Status: Offline Project Badges: |
I am not sure what has changed in the past few weeks (or months). But something needs to be changed back I think. I do not monitor BOINC all the time, but have recently noticed that work units list show double or triple the time to complete over what they have in the past. That is to say they show a time to complete of over a day each (and they do take about that long). That would be okay for me, but I see this thread saying 18 hour limit and perhaps this explains why I see so many results being "ERROR" after running twenty hours or more and resulting in no credit for all the work. Is the problem in BOINC or the project itself? I was looking to get a ten year badge but I may need to cut my loss and move on...
|
||
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 1978 Status: Offline Project Badges: |
I see this thread saying 18 hour limit and perhaps this explains why I see so many results being "ERROR" after running twenty hours or more and resulting in no credit for all the work. So far I've never seen not getting any credit for running 18 CPU hours ending in Error. See e.g. https://secure.worldcommunitygrid.org/forums/wcg/viewpostinthread?post=512391 and then - the results, now wíth credit - 2 posts later: https://secure.worldcommunitygrid.org/forums/wcg/viewpostinthread?post=512419. There's also an explanation by SekeRob*: https://secure.worldcommunitygrid.org/forums/wcg/viewpostinthread?post=509704. [Edit 3 times, last edit by adriverhoef at Mar 10, 2016 11:03:31 AM] |
||
|
Procyon Lotor One
Cruncher Joined: Jun 28, 2008 Post Count: 14 Status: Offline Project Badges: |
Thanks adriverhoef. That was an interesting read.
True, it is hard for me to tell in retrospect if it showed no credit with an error. I had moved the project from a PC to a server to see what would happen - with about the same results. My overall impression was that the CPU was working all day for 60 or 70 credits per WU. I probably should have looked deeper into the issue. I ended up moving the sever to another project for now. |
||
|
Sandvika
Advanced Cruncher United Kingdom Joined: Apr 27, 2007 Post Count: 112 Status: Offline Project Badges: |
I routinely get zero credit for error work units. I limped to Gold and then quit, now with HST and FAH2 both having really long WUs too I figured I'd have another go. No change, ruby seems elusive. With so much crunching done for no useful work I'm edging towards the exit again. I'd like to get 5 years on any current project as a minimum but this one being 32-bit only and time limited per WU makes poor use of the available hardware.
---------------------------------------- |
||
|
|