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: 387
|
![]() |
Author |
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1114 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() |
Great update. They have been working hard ! Glad they mentioned working on ARP next
|
||
|
adict2jane
Cruncher Joined: Aug 18, 2006 Post Count: 32 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Thank you for the update!
|
||
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12559 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
20/04/2025 13:07:55 | World Community Grid | Not requesting tasks: don't need (CPU: max concurrent job limit; Intel GPU: blocked by project preferences)
This is despite neither ARP nor MCM being at their cache limits. Mike |
||
|
gj82854
Advanced Cruncher Joined: Sep 26, 2022 Post Count: 120 Status: Offline Project Badges: ![]() ![]() |
That was a fix that was applied a while back to the client as a result of the client ignoring work fetch limits when max concurrent was specified in the app_config.xml file.
|
||
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1058 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
That was a fix that was applied a while back to the client as a result of the client ignoring work fetch limits when max concurrent was specified in the app_config.xml file. The fix uses the smallest max_concurrent value for a specific [BOINC] project for its decision making because there's no way the client can tell a server not to send work for a particular application... This can cause issues if the client has mis-calculated the estimated run time for one or more of the applications[*1]; if the estimates are excessive the only work-around seems to be to increase the cache size!If one is only running BOINC projects that support server-side limits on tasks sent, this isn't a problem[*2]. However, not all projects do so, and in those cases that could result in other projects flooding the cache :-( I don't think that fix is likely to be changed, so if Mike is only running WCG work I'd suggest he makes sure his WCG profile limits are such that he can never get more work than he wants and then increase buffer size -- on all but one of my systems I have needed to do something like that, with the buffer size now having at least twice the space needed for the maximum number of tasks it will ever receive! Cheers - Al. *1 I have a self-compiled 8.0.2 client (to solve the issue of non-default venues causing log spam issues, a problem mentioned elsewhere , and first self-resolved for client 7.24.1) -- when building 8.0.2 I may have missed a compilation option that has affected the benchmarks, which are not consistent with those shown by an unpatched client... *2 Sites with standard BOINC and profile limits may report that a profile-based limit has been reached on each unnecessary request; WGC doesn't bother to report that! |
||
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1114 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() |
Thanks Al
I've noticed the weirdness and was trying to figure out what was happening. I did not get the nice debug message Mike had, so I was lost. |
||
|
hchc
Veteran Cruncher USA Joined: Aug 15, 2006 Post Count: 837 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Happy Easter!
----------------------------------------I've gotten tons of ARP1 tasks the past week or so, more than I've received in many months.
|
||
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1114 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() |
My flow of ARP WUs is dwindling. I got a resend at 2025-04-20 23:56:52 UTC but it has been longer since I got a fresh ARP WU.
----------------------------------------MCM is still flowing well. edit: now getting fresh ARPs edit again: small window of fresh ARPs. not getting any again. [Edit 2 times, last edit by Unixchick at Apr 22, 2025 12:34:52 AM] |
||
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2277 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Getting all the tasks I want, and all I want is MCM, while waiting for MAM Beta for Windows.
|
||
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1114 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() |
The reference number leads me to believe that ARP WUs are going out at an OK rate. I'm not getting many personally though.
MCM is flowing well. I'm not getting one every ask, but good enough. |
||
|
|
![]() |