| 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: 8
|
|
| Author |
|
|
mariokonrad
Cruncher Joined: Dec 8, 2008 Post Count: 5 Status: Offline Project Badges:
|
Hey,
sometimes in the "Result Status" section I find a WU with "too late". Some of my friends help me crunching with their devices, but a lot of them use their computer only about 1 hour a day and most of the big FightAids@home WU's are finished too late. CPU TIME: 19.04 157.5 / 0.0 I get really sad when I see something like this. I'd suggest to give people more time, like one month. Don't get me wrong, I don't want to download more WU's (10 days is enough) I just want more time for my friends to finish such big WU's. All WU's that get finished to late, are a waste of energy. A bit more time to finish WU's could make wcg more effective, because only bad WU's would be aborted. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Why should I wait a month for a 'partner' in the replication to finish? no thank you. I would prefer the time to be 3 days, much better!
![]() |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges:
|
Even if the computer is on 1 hr/day, then a FA@H WU should be returned in time - as they generally run for ~ 8hrs.
----------------------------------------Perhaps there's something in their settings that's stopping BOINC from running in the background? ![]() |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
I'd suggest to give people more time, like one month. Hi mariokonrad!... I just want more time for my friends to finish such big WU's. This is not realistic, not only because of some brutal members like Scribe, but more because that would dramatically slow down the whole process and excessively increase the number of jobs "in progress" in the main database. But "we have what you need"! HCMD2 jobs have time limitations as follows: 1. if a job has not reached 60 % of the work to do after 6 hours of CPU time the job is stopped at the end of the current computation loop and the work not done will be redistributed in another job. 2. if the job has done more than 60 % after 6 hours but still does not complete after 12 hours, then it is stopped the same. In most cases machines with the usage profile you describe will return HCMD2 tasks after 6 hours, and the few which might fall in category two would still return in time if they are switched on only one hour and 12 mn per day instead of one hour. Cheers. Jean. PS: We still have another active project (NRW/Rice) which is limited to 7 hours in all cases, but it is on its end and there are no new tasks on distribution. ---------------------------------------- [Edit 1 times, last edit by JmBoullier at Apr 5, 2010 3:37:42 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I'd suggest to give people more time, like one month. Don't get me wrong, I don't want to download more WU's (10 days is enough) 10 days buffer is way too big for a machine that is not on 24x7. Cut it back to 1 day at most or 2 days if the machine manually connects to the Internet. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
My buffer settings are .5 days and cache of .2......
----------------------------------------![]() ....and on 365/24/7 [Edit 1 times, last edit by Former Member at Apr 5, 2010 4:22:16 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
A month! Egads man...
I thing there should be only a 3 or 4 day buffer allowed here. There is plenty of work here at WGC to be worried about running out of work. I have mine set to 0.1 and 0.1 running 24/7/365 ![]() |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
No need to get nervous, wise cachers: the BOINC client quickly learns about its usage profile, and even with a 10-day cache (which is too much for me too) probably it would not load the next task much before the end of the current one. Assuming it would even "take the risk" before being sure that the current one can be returned in time.
----------------------------------------If Ingleside reads this thread he will probably give you the maximum settings that BOINC would use to have a chance to manage these extreme cases successfully. Cheers. Jean. |
||
|
|
|