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: 93
|
![]() |
Author |
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
We are currently implementing a scheduling change that allows only 1 work unit per host for this project. This is being done to accomplish a few things. The main being disk IO on your computer. From our alpha and beta tests, we have seen computer usability issues due to the large disk IO. Limiting the project to one per host will help keep this from happening.
Thanks, -Uplinger |
||
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Frankly, on Linux not seen usability issues... just disproportional time loss that is used by BOINC but not recorded as BOINC time.
----------------------------------------Now it's up to you guys to keep that feeder prioritizing these CEP2 jobs so that when one is finished, a new one is provided. Given that the client backfills the buffer before a task finishes the effect is that from before about 6-7 on my quad, there will practically be 2, tops 3 daily and I think this is an optimistic view. PS, There's a cruncher who apparently made a special build of 6.10.58 that can control the number of concurrent tasks of same science, read on alpha mail list. Something to research. cheers edit: remove name.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Oct 23, 2010 5:22:35 AM] |
||
|
Dataman
Ace Cruncher Joined: Nov 16, 2004 Post Count: 4865 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
The one-per-host part is certainly working.
----------------------------------------![]() ![]() |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Sekerob may have made this statement in his own words. I hope that IF I have CEP checked, when a system requests a new WU and I am currently not running a CEP WU, the feeder will give me a CEP WU first, unless there are no CEP WU in the feeder. AND that is does this all the time. AND hopefully the feeder never runs dry of CEP WU. This only will apply if the system in question meets the CEP minimums.
|
||
|
X-Files 27
Senior Cruncher Canada Joined: May 21, 2007 Post Count: 391 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
I've only seen usability issues when running linux on usb stick other than that its smooth sailing.
----------------------------------------This scheduling change is more like a workaround rather than a fix. ![]() ![]() |
||
|
Jack007
Master Cruncher CANADA Joined: Feb 25, 2005 Post Count: 1604 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
hmmm well I guess this means i'll have my slow
----------------------------------------computers working on the windows ver when it comes out. (well my fast ones can catch one here or there that's good too) It'll be fun to see how long a 1.6 ghz machine takes (if it even qualifies to get one) ![]() |
||
|
X-Files 27
Senior Cruncher Canada Joined: May 21, 2007 Post Count: 391 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
can we get an exception for those computers that would like to crunch on CEP2 only?
----------------------------------------![]() ![]() |
||
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 2986 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Also, will these CEP2 WU's still have the 10 day deadline, or a shorter one? (i.e., so that they'll jump to the front of the queue - therefore finishing quicker and allowing another to be downloaded straightaway)
----------------------------------------![]() |
||
|
codes
Advanced Cruncher Joined: Oct 20, 2009 Post Count: 142 Status: Offline |
1 per host has worked well for my single core 1.8 GHz Celeron running Slackware 13.1 x64 with x86 libs. It's running strickly CEP2 for the past two and half months and I haven't seen any of the low CPU times other than the normal 30 - 45 minute wall clock to CPU time difference at the end. My CPU time is on average 12 hours per WU. I'm still running the orginal 2.6.33.4 kernel that came with 13.1, but kinda worried if I upgrade to a newer one I might start seeing problems like other CEP2 crunchers have reported lately.
----------------------------------------Edit: added additional comment. [Edit 1 times, last edit by codes at Oct 22, 2010 9:58:21 PM] |
||
|
dango
Senior Cruncher Joined: Jul 27, 2009 Post Count: 307 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
it's not so good idea, to limit crunching. my quad (4gb ram-gentoo) has not big IO issue
|
||
|
|
![]() |