| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Community
Forum: Chat Room
Thread: Project Status (First Post Updated) |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 260
|
|
| Author |
|
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1295 Status: Offline Project Badges: 
|
Time to pull out the backup projects ! I got one resend today. It is a holiday weekend so don't expect a fix until Monday.
|
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2497 Status: Recently Active Project Badges: 
|
And now we're also back to this problem:
----------------------------------------503 Service Unavailable No server is available to handle this request. And the production of work is falling badly: MCM1 03/30/2024 222,012 Results returned (midday today, 3/30/24 12:06:04 UTC) OPNG 03/30/2024 27,375 Results returned (midday today, 3/30/24 12:06:04 UTC) [Edit 3 times, last edit by Grumpy Swede at Mar 30, 2024 7:40:18 PM] |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7846 Status: Offline Project Badges: 
|
And now we're also back to this problem: 503 Service Unavailable No server is available to handle this request. And the production of work is falling badly: MCM1 03/30/2024 222,012 Results returned (midday today, 3/30/24 12:06:04 UTC) OPNG 03/30/2024 27,375 Results returned (midday today, 3/30/24 12:06:04 UTC) I think the number of work units being returned is the result of those with large caches eventually getting worked down. Since the techs do not work on the weekends, I would speculate Sunday's totals will be even more anemic. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1317 Status: Offline Project Badges: 
|
From about 06:45 UTC on 2024-03-31 I started receiving new MCM1 tasks; as at 07:20 UTC all my systems have their normal, small, full quota of work :-)
----------------------------------------OPNG work also still seems to be flowing, but not quite as much as a couple of days ago... Cheers - Al P.S. the "Error 503" issue is still showing up quite a lot on the forums and, occasionally, on the results pages. [Edit 1 times, last edit by alanb1951 at Mar 31, 2024 7:23:52 AM] |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2497 Status: Recently Active Project Badges:
|
9 times out of 10, I get "Error 503", or "System Error" and it takes forever to get anything posted here. To post this, I had to try for 15 minutes. The results page is almost impossible to get to load fully.
----------------------------------------A few MCM tasks comes in, but a long time in between, and only _2 resends. No OPNG's at all for many hours. But it's OK really. I'm alive and reasonably healthy  [Edit 1 times, last edit by Grumpy Swede at Mar 31, 2024 9:36:55 AM] |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2497 Status: Recently Active Project Badges:
|
Much better flow of both MCM and OPNG tasks now. The website also seems to work much better, at least for now.
----------------------------------------[Edit 1 times, last edit by Grumpy Swede at Mar 31, 2024 5:51:01 PM] |
||
|
|
TLD
Veteran Cruncher USA Joined: Jul 22, 2005 Post Count: 856 Status: Offline Project Badges: 
|
Much better flow of both MCM and OPNG tasks now. The website also seems to work much better, at least for now. Same here, I wonder if everybody turned on their backup projects leaving more WUs for the folks still here.  |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1317 Status: Offline Project Badges:
|
Much better flow of both MCM and OPNG tasks now. The website also seems to work much better, at least for now. Same here, I wonder if everybody turned on their backup projects leaving more WUs for the folks still here. I seem to recall mention of something WCG run to try to regulate the work supply, and I wonder how well its latest manifestation copes with situations where most of the work out there will take three or more days to be returned because of users increasing cache sizes to cope with work shortages. The new work flow started after about 24 hours of getting nothing but retries because of missed deadlines; if I can believe the API wingman feed I have organized, almost all the retries I received were related to a fairly small group of systems which usually seem to return stuff after 4 days or longer when not missing deadlines (so there are some systems with large buffers or other reasons for late returns ...) If that has been happening, it might also explain why the tech team thought there was plenty of work available -- work out in the field isn't necessarily evenly distributed between those who need it to keep cores busy and those who want it because they're worried about running out :-) Cheers - Al. P.S. Even when work flow is more even, a significant proportion of the non-retry MCM1 work I return sits waiting for 4 or more days before a wingman returns, and it's rare to have everything validate without any missed deadline retries coming into play :-( -- and this was just as true when twice as many results a day were being returned... |
||
|
|
Paul Schlaffer
Senior Cruncher USA Joined: Jun 12, 2005 Post Count: 278 Status: Offline Project Badges: 
|
On the bright side I came back from my weekend travels to see my workstation crunching WCG, both MCM & OPNG, back from the dead.
---------------------------------------- Seems appropriate for Easter Sunday. Have a good Easter everyone. Much better flow of both MCM and OPNG tasks now. The website also seems to work much better, at least for now. Same here, I wonder if everybody turned on their backup projects leaving more WUs for the folks still here. I seem to recall mention of something WCG run to try to regulate the work supply, and I wonder how well its latest manifestation copes with situations where most of the work out there will take three or more days to be returned because of users increasing cache sizes to cope with work shortages. The new work flow started after about 24 hours of getting nothing but retries because of missed deadlines; if I can believe the API wingman feed I have organized, almost all the retries I received were related to a fairly small group of systems which usually seem to return stuff after 4 days or longer when not missing deadlines (so there are some systems with large buffers or other reasons for late returns ...) If that has been happening, it might also explain why the tech team thought there was plenty of work available -- work out in the field isn't necessarily evenly distributed between those who need it to keep cores busy and those who want it because they're worried about running out :-) Cheers - Al. P.S. Even when work flow is more even, a significant proportion of the non-retry MCM1 work I return sits waiting for 4 or more days before a wingman returns, and it's rare to have everything validate without any missed deadline retries coming into play :-( -- and this was just as true when twice as many results a day were being returned... If people are hording days of work, yes that's a problem.
“Where an excess of power prevails, property of no sort is duly respected. No man is safe in his opinions, his person, his faculties, or his possessions.” – James Madison (1792)
----------------------------------------[Edit 1 times, last edit by Paul Schlaffer at Mar 31, 2024 11:19:56 PM] |
||
|
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1295 Status: Offline Project Badges:
|
Glad to hear the WU flow is better. Thank you to any tech who took time to fix things. I still haven't seen a fresh MCM WU yet, but I am doing manual requests to try and request more frequently.
|
||
|
|
|