| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Community
Forum: Chat Room
Thread: Project Status (old) |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 387
|
|
| Author |
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1317 Status: Offline Project Badges: 
|
Given that MAM is based on MCM, does this imply that all of the work will moving to GPU in the near term? I'll be surprised if they take MCM1 or MAM1 to GPU only, especially if they are going to target NVIDIA...I will not run a big GPU and my GT710s are not exactly powerful :-) If GPU tasks will take a reasonable time to run, they might be able to work in roughly the same way MilkyWay Separation did, with WUs going to GPUs or CPUs depending on demand (though, unlike MW Separation, they would probably do the strict hardware and O/S matching...). If the algorithms for CPU and GPU are fundamentally the same they may not need to separate them out in the way they did with OPN1 and OPNG (which used totally different algorithms!), though it might make assigning credit quite interesting as it can't sensibly be based on run-time when a GPU is involved! That said, if GPUs can do the work an order of magnitude faster than CPUs, it might be that separate streams would be needed so that they could either batch up work for GPUs (as per OPNG) or give "harder" work to GPUs. It'll be interesting to see how this pans out! As for older GPUs... I suspect they'll only run work on more recent GPUs with higher CUDA compute capabilities; I guess my 1650Ti might be o.k. but my 1050Ti might be too old... Cheers - Al. [Edit 1 times, last edit by alanb1951 at May 27, 2025 4:07:28 AM] |
||
|
|
hchc
Veteran Cruncher USA Joined: Aug 15, 2006 Post Count: 865 Status: Offline Project Badges: 
|
That's cool they're prioritizing the Kubernetes migration, so hopefully we get a bit more stability soon.
----------------------------------------I'm also excited for MAM1 to become ready for production. To save heat/electricity, I turned off MCM1 and am doing ARP1 only. Mostly idle, but every now and then some tasks trickle in. Only 1 in the past 2-3 days.
|
||
|
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1293 Status: Offline Project Badges: 
|
I hope they are working on something as I just got this in my logfile
----------------------------------------World Community Grid | Server can't open database World Community Grid | Project requested delay of 3600 seconds The fact that the requested delay is longer than usual, makes me think this is a planned outage. edit: error message now "feeder not running" edit: no longer getting error messages. hopefully will get some WUs soon. [Edit 3 times, last edit by Unixchick at May 29, 2025 10:00:03 PM] |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1317 Status: Offline Project Badges:
|
[Edit: I composed and posted this before I saw that Unixchick's report had been edited to reflect system recovery. I'll leave it more or less "as is"...]
----------------------------------------My script that samples the ARP1 Project Statistics History page had intermittent problems from 08:50 to 11:10 (UTC) on 2025-05-29, and the forums were slow to shift from page to page (internal networking issue?). However, all my other API stuff was working until about 19:50 UTC. Status changed about 2 hours later... 2025-05-29 21:30 UTC -- Server can't open database 2025-05-29 21:40 UTC -- Server error: feeder not running And at 21:50 UTC one of my systems successfully reported 32 tasks, though it was another few minutes before a request got some new work to replace some of those! This looks as if it was an internal network issue rather than a database problem per se. The eventual recovery seems too swift to have Involved a complete system restart :-) Here's hoping that that's it for a few days now! Cheers - Al. P.S. I think the 1 hour delay request is standard for certain server conditions, but a code dive (or a BOINC server guru's say so) would be needed to confirm that. [Edit 2 times, last edit by alanb1951 at May 29, 2025 10:10:50 PM] |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges:
|
Reporting now available but tasks committed to other platforms
Mike |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges:
|
MCM downloading.
Mike |
||
|
|
terrycmora
Cruncher Joined: Jan 31, 2014 Post Count: 1 Status: Offline Project Badges: 
|
I just wanted to say a quick thank you to everyone involved. Putting you time and energy out for the good of everyone. It is thankless and rarly seen by most. But it is heartfelt.
|
||
|
|
TLD
Veteran Cruncher USA Joined: Jul 22, 2005 Post Count: 856 Status: Offline Project Badges:
|
Requesting new tasks for CPU and NVIDIA GPU
----------------------------------------Scheduler request completed: got 0 new tasks Must be the weekend.  |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges:
|
TLD
Firstly, GPU units last went out months ago. Secondly, ARP & MCM are going out but spasmodically. You need luck to request at the right moment. Mike |
||
|
|
Unixchick
Veteran Cruncher Joined: Apr 16, 2020 Post Count: 1293 Status: Offline Project Badges:
|
I think ARP fresh WUs have stopped. The reference number is falling. Anyone got a fresh one lately?
|
||
|
|
|