| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: File System Error Update thread |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 62
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Is it possible to have an update on the problem please? I shall run out of work in about 4 hours and it would be nice to have a forecast on possible completion.
Thanks |
||
|
|
X-Files 27
Senior Cruncher Canada Joined: May 21, 2007 Post Count: 391 Status: Offline Project Badges: 
|
Yes, update pls.
----------------------------------------Some of my rigs are out of work. Maybe its time to increase the cache size to 1 day from .5?   [Edit 1 times, last edit by X-Files 27 at Oct 14, 2010 8:14:33 AM] |
||
|
|
Hypernova
Master Cruncher Audaces Fortuna Juvat ! Vaud - Switzerland Joined: Dec 16, 2008 Post Count: 1908 Status: Offline Project Badges: 
|
It is amazing that I decided a few days ago to increase the cache in all my devices from 1 day to 3 days. Hope it will be enough.
----------------------------------------I must have some kind of "mentalist" capabilities.  Next time I have a bad feeling I let you know.   |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Regular "Set / Forget" crunchers do best to ignore most of these discussions of 0.1 cache sizes. These are with an some agenda such as be able to max out on fetching any rain/deluge of DDDT2 Type B/C, and for catching Beta/DDDT2 Type A it anyway wont work since those are since a while limited to 1 per core.
----------------------------------------The WCG device profile defaults are/were 0.3 days for "additional buffer" and 0.2 for "Connect Every..." which totals to an effective minimum cache of 0.5 days or 12 hours, meaning if a client has even 1 second less than 12 hours it goes to fetch more work if the network is open. Personally, as Thé Machine grows towards 0.5 PetaFLOPS, the probability notwithstanding all the safeties and checks and redundancies are, that things such as file-system outages will take longer to verify. In that, my personal recommendation is around 1 day buffer (the sum of connect+additional buffer). That will cover 99.9% of the outages. 2 days buffer is definitely save. Except one long ago X-mas, not ever seen such downtimes at WCG. v.v. Scribe's question "how long"... pass. There was the GPFS change over some days ago, so it's not unlikely that more tweaking will happen with this experience. The bigger the SAN/RAID storage the longer it will take as each file has to be matched up with the central registries. edit: thís forum software continues to have serious issue in taking in any special character.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Oct 14, 2010 9:21:14 AM] |
||
|
|
GB033533
Senior Cruncher UK Joined: Dec 8, 2004 Post Count: 206 Status: Offline Project Badges: 
|
Well, that's me all finished. All wus in status 'uploading', except one 'ready to report'.
----------------------------------------Just waiting and hoping.  |
||
|
|
guenterhb
Cruncher Joined: Sep 22, 2006 Post Count: 10 Status: Offline Project Badges: 
|
well, if somebody (like me) runs out of steam with WCG work, its time to try-out other BOINC projects temporarily - better than idling the CPUs
 |
||
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: 
|
It is amazing that I decided a few days ago to increase the cache in all my devices from 1 day to 3 days. Hope it will be enough. I must have some kind of "mentalist" capabilities. Next time I have a bad feeling I let you know. Yes - send me an email so I can clear my schedule.... |
||
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges:
|
Thanks so much for the the update, Kevin. We appreciate all your hard work!
---------------------------------------- |
||
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges:
|
well, if somebody (like me) runs out of steam with WCG work, its time to try-out other BOINC projects temporarily - better than idling the CPUs Good point! Does anyone have suggestions of good projects to try -- with short-running WUs so they'll finish quickly and we can get right back to WCG as soon as it's back online? :) |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
well we did a fsck last month 'cause of an error on our CAD/CAM server.....
there were about 4 TB HD's and it tooks plenty time to check, i think there was about 16 hours to complete......  I have no idea how big is the fs of WCG but i think that will take about 20 to 30 hours to complete the check  the problem was that i could not connect to WCG during this time so i set the buffer on 5 days on every pc i have  but what's about the results who were outrunning (return time 10 days) ? will they give some points or not ?  (yes i know, we crunch for helping and not for points........  )greez Marcel |
||
|
|
|