Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
World Community Grid Forums
Category: Community Forum: Chat Room Thread: Project Status Information (first post updated) |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 155
|
Author |
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 859 Status: Offline Project Badges: |
I can see that MCM is flowing at a great rate now. Unfortunately it isn't enough.
I'm running MCM one at a time, just to keep my machine asking for other tasks, and I can't even get my one MCM reliably . |
||
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7581 Status: Recently Active Project Badges: |
I can see that MCM is flowing at a great rate now. Unfortunately it isn't enough. I'm running MCM one at a time, just to keep my machine asking for other tasks, and I can't even get my one MCM reliably . Two reasons the flow of MCM can not keep up with demand are 1) It is currrently the only project running and 2) The average run time per unit in the last couple of days appears to be about 80% of the previous work units which means machines are requesting work units much more frequently than in the past. Cheers
Sgt. Joe
----------------------------------------*Minnesota Crunchers* [Edit 1 times, last edit by Sgt.Joe at Sep 10, 2023 4:14:34 PM] |
||
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 873 Status: Offline Project Badges: |
I can see that MCM is flowing at a great rate now. Unfortunately it isn't enough. I'm running MCM one at a time, just to keep my machine asking for other tasks, and I can't even get my one MCM reliably . Two reasons the flow of MCM can not keep up with demand are 1) It is currrently the only project running and 2) The average run time per unit in the last couple of days appears to be about 80% of the previous work units which means machines are requesting work units much more frequently than in the past. Cheers Cheers - Al. [Edit 1 times, last edit by alanb1951 at Sep 10, 2023 6:34:03 PM] |
||
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 859 Status: Offline Project Badges: |
I haven't gotten any WUs in hours.
|
||
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 873 Status: Offline Project Badges: |
I wondered about intermittent supply, so I've just had a look through the boinc.log files on the 4 systems that try to get work from WCG, looking for "Tasks are committed to other platforms" messages during 11th September (BST [UTC+1]). These are all Linux systems, by the way...
----------------------------------------Based on what I saw, there still seems to be a [tiny?] set of retries for SCC1, but I suspect they may all be for Android :-) More to the point re MCM1, I found quite a few occasions where the "other platforms" message was present, SCC1 reported no tasks available and MCM1 was missing -- perhaps we've reached another "here come lots of missed-deadline retries" period, in which case that might explain apparent slowing of available work. Unfortunately, I suspect Mac users might get the worst of that :-( I saw quite a few MCM1 "other platforms" incidents between 07:00 and 20:00 (UTC) , with a much smaller number outside that period. I have quotas set in my device profiles, and I think it will report "No tasks" when my quota is met, even if tasks are committed to other platforms, so my observations may be under-reporting... As it happens, I currently have no problems keeping my [small] quotas topped up, although it has sometimes taken a couple of hours for a completed, reported task to be replaced. However, I note that Sgt. Joe has just posted about [most of] his Linux boxes having run dry, so I also wonder if there's an option in play giving preference to small requests... Cheers - Al. [Edit - revised the top-up comment (and [Edit 4 times, last edit by alanb1951 at Sep 14, 2023 11:20:12 PM] |
||
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 859 Status: Offline Project Badges: |
The forum is back. Thanks Tech people.
Looks like MCM and some OPG WUs are going out. |
||
|
Hans Sveen
Veteran Cruncher Norge Joined: Feb 18, 2008 Post Count: 802 Status: Offline Project Badges: |
Back up?`!
----------------------------------------MCM units have been availible while the site was down, even some OPNG units came along 😍 PS. Just ignore, Unixchick typed faster than old me!! [Edit 2 times, last edit by Hans Sveen at Sep 14, 2023 7:58:05 PM] |
||
|
thunder7
Senior Cruncher Netherlands Joined: Mar 6, 2013 Post Count: 232 Status: Offline Project Badges: |
well, MCM is available, but not in quantities like before. My cpu's are not dry, but they seem to hover much closer to it than say 10 days ago.
48 cores, 48 tasks running, 19 tasks waiting 40 cores, 40 tasks running, 15 tasks waiting I used to see about 950 tasks waiting, which was a comfortable buffer, given the frequency of weekend problems. It's friday again, and buffers are low. What gives? |
||
|
bfmorse
Senior Cruncher US Joined: Jul 26, 2009 Post Count: 294 Status: Offline Project Badges: |
Based on the Sent/Due/Returned Dates in the component WU's, sadly it appears that some individuals are still running six day queues.
To maximize my chance of successfully completing those WU's assigned to me before they get aborted, I run a queue of ZERO DAYS and am mostly successful in getting things processed in time. |
||
|
thunder7
Senior Cruncher Netherlands Joined: Mar 6, 2013 Post Count: 232 Status: Offline Project Badges: |
Even if it were a conscious decision to override buffer settings on clients, it'd be nice if there was some communication about that.
----------------------------------------[Edit 1 times, last edit by thunder7 at Sep 15, 2023 6:37:38 AM] |
||
|
|