SETI@Home will now cache 150 CPU work units and 150 per installed GPU

Message boards : Number crunching : SETI@Home will now cache 150 CPU work units and 150 per installed GPU
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · Next

AuthorMessage
Lazydude
Volunteer tester

Send message
Joined: 17 Jan 01
Posts: 45
Credit: 96,158,001
RAC: 136
Sweden
Message 2022068 - Posted: 7 Dec 2019, 11:22:52 UTC
Last modified: 7 Dec 2019, 11:24:50 UTC

Hi!
I have only one GPU

12/07/19 12:01:56 | SETI@home | Sending scheduler request: To fetch work.
12/07/19 12:01:56 | SETI@home | Reporting 3 completed tasks
12/07/19 12:01:56 | SETI@home | Requesting new tasks for NVIDIA GPU
12/07/19 12:02:00 | SETI@home | Scheduler request completed: got 148 new tasks

looks like the limit of 100 wus per device is lifted or something else have happend
at the moment i have 117 wu for CPU
and 230 wu for GPU
ID: 2022068 · Report as offensive     Reply Quote
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14660
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2022069 - Posted: 7 Dec 2019, 11:23:25 UTC - in response to Message 2022064.  

Are you saying that the powers-that-be undid the 100 max WUs?
I'm not - I keep my cache sizes low anyway, so it barely affects me. But that's what everyone else seems to be saying.

Even if true, we don't yet know whether the change was deliberate/permanent, or a mistake/temporary. Time will tell.
ID: 2022069 · Report as offensive     Reply Quote
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 2022074 - Posted: 7 Dec 2019, 11:56:15 UTC - in response to Message 2022069.  

Are you saying that the powers-that-be undid the 100 max WUs?
I'm not - I keep my cache sizes low anyway, so it barely affects me. But that's what everyone else seems to be saying.

Even if true, we don't yet know whether the change was deliberate/permanent, or a mistake/temporary. Time will tell.



I seem to be at a bottleneck for uploads and downloads at my site. Slowly still getting more downloaded than going back so the caches are building. Won't know yet for a few hours but I'm thinking they raised the limit to 400 work units per GPU? Anyone else seeing that?
ID: 2022074 · Report as offensive     Reply Quote
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14660
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2022075 - Posted: 7 Dec 2019, 12:01:45 UTC - in response to Message 2022074.  

I'm thinking they raised the limit to 400 work units per GPU? Anyone else seeing that?
I'm not seeing that, but then I wouldn't even begin to look until the system has recovered somewhat.

The only way to know is to wait until you see a line about 'reached a limit on tasks in progress' in the Event Log, and then go back and count then. I might try that come about Monday, if things have settled down, but not now.
ID: 2022075 · Report as offensive     Reply Quote
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13797
Credit: 208,696,464
RAC: 304
Australia
Message 2022076 - Posted: 7 Dec 2019, 12:02:45 UTC - in response to Message 2022074.  

Won't know yet for a few hours but I'm thinking they raised the limit to 400 work units per GPU? Anyone else seeing that?
No way of telling at this stage for me. Had to reduce my cache settings to stop my Linux system from running out of CPU work as it's till struggling to get anything.
Grant
Darwin NT
ID: 2022076 · Report as offensive     Reply Quote
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14660
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2022077 - Posted: 7 Dec 2019, 12:20:00 UTC

OK, can't resist. I have one machine running the dynamic spoofed client. It has two GPUs, and (currently) 948 tasks - so it should be just above the 'Zalster limit'. I've set it back to spoof 2 GPUs, and we'll see what happens. So far - "no tasks available", no reason given.

Or perhaps - bingo.

07/12/2019 12:17:47 | SETI@home | This computer has reached a limit on tasks in progress
ID: 2022077 · Report as offensive     Reply Quote
Profile Tom M
Volunteer tester

Send message
Joined: 28 Nov 02
Posts: 5124
Credit: 276,046,078
RAC: 462
Message 2022081 - Posted: 7 Dec 2019, 12:47:29 UTC

In another thread. It was reported that the "limits would be increased slowly" so that high CPU count systems (like a 128 thread Threadripper or an Eypc of similar size) would get enough work. It wasn't clear from my reading that this would also change the gpu limits.
However, they did say "slowly" and apparently it is not so slowly. Plus we have the inconsistent access issue that everyone with multiple machines is reporting.

Tom
A proud member of the OFA (Old Farts Association).
ID: 2022081 · Report as offensive     Reply Quote
juan BFP Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 16 Mar 07
Posts: 9786
Credit: 572,710,851
RAC: 3,799
Panama
Message 2022083 - Posted: 7 Dec 2019, 12:55:28 UTC - in response to Message 2022077.  
Last modified: 7 Dec 2019, 12:56:56 UTC

OK, can't resist. I have one machine running the dynamic spoofed client. It has two GPUs, and (currently) 948 tasks - so it should be just above the 'Zalster limit'. I've set it back to spoof 2 GPUs, and we'll see what happens. So far - "no tasks available", no reason given.

Or perhaps - bingo.

07/12/2019 12:17:47 | SETI@home | This computer has reached a limit on tasks in progress

I running the unlimited spoofed client as you all know.
The only problem is to get any new work the rest is working as usual.
Just get 108 new task from a request.
Sat 07 Dec 2019 07:47:14 AM EST | SETI@home | Requesting new tasks for CPU and NVIDIA GPU
Sat 07 Dec 2019 07:47:18 AM EST | SETI@home | Scheduler request completed: got 108 new tasks

What i believe is happening is: the limit was changed/brooked and something is not working fine, or simply as we could expect the servers are having a hard time to adapt to this new limits. The time will tell.
ID: 2022083 · Report as offensive     Reply Quote
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14660
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2022089 - Posted: 7 Dec 2019, 13:09:01 UTC - in response to Message 2022081.  

In another thread. It was reported that the "limits would be increased slowly" ...
A full forum search doesn't turn up that phrase, at least in the last 6 months. Who reported, when and where? Was it a project admin, or just wishful thinking by a volunteer?
ID: 2022089 · Report as offensive     Reply Quote
Profile Bernie Vine
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 26 May 99
Posts: 9954
Credit: 103,452,613
RAC: 328
United Kingdom
Message 2022091 - Posted: 7 Dec 2019, 13:14:16 UTC - in response to Message 2022089.  

In another thread. It was reported that the "limits would be increased slowly" ...
A full forum search doesn't turn up that phrase, at least in the last 6 months. Who reported, when and where? Was it a project admin, or just wishful thinking by a volunteer?


This is the post in question


https://setiathome.berkeley.edu/forum_thread.php?id=83824&postid=2022034
ID: 2022091 · Report as offensive     Reply Quote
Profile Tom M
Volunteer tester

Send message
Joined: 28 Nov 02
Posts: 5124
Credit: 276,046,078
RAC: 462
Message 2022092 - Posted: 7 Dec 2019, 13:14:18 UTC - in response to Message 2022089.  

In another thread. It was reported that the "limits would be increased slowly" ...
A full forum search doesn't turn up that phrase, at least in the last 6 months. Who reported, when and where? Was it a project admin, or just wishful thinking by a volunteer?


Take a look at https://setiathome.berkeley.edu/forum_thread.php?id=83824&postid=2022042#2022042

Tom
A proud member of the OFA (Old Farts Association).
ID: 2022092 · Report as offensive     Reply Quote
Profile Mr. Kevvy Crowdfunding Project Donor*Special Project $250 donor
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 15 May 99
Posts: 3783
Credit: 1,114,826,392
RAC: 3,319
Canada
Message 2022093 - Posted: 7 Dec 2019, 13:17:18 UTC

I wrote to Dr. Korpela (also re: the AMD RX 5700 issue) but of course it's five in the morning on a Saturday in Cali so I hope his rest is undisturbed for a few hours. :^)

Just before this started I noticed there was a very brief site outage with a malformed maintenance notification in error code, so it would appear that changes were made. Hopeful there will be confirmation.
ID: 2022093 · Report as offensive     Reply Quote
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14660
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2022099 - Posted: 7 Dec 2019, 13:37:27 UTC - in response to Message 2022091.  

This is the post in question

https://setiathome.berkeley.edu/forum_thread.php?id=83824&postid=2022034
Thanks. And Github seems to confirm that captainiom (with whom I'd already corresponded) is the same person as JSM. We're beginning to get an idea of the picture.
ID: 2022099 · Report as offensive     Reply Quote
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14660
Credit: 200,643,578
RAC: 874
United Kingdom
Message 2022101 - Posted: 7 Dec 2019, 13:42:52 UTC - in response to Message 2022077.  

Well, I'm down below 800 tasks on my 2 x GPU machine, and I'm back to 'no work available', with no mention of a limit. Looks like "Zalster's Theorem" is right, but I won't know for certain until I get back from lunch.
ID: 2022101 · Report as offensive     Reply Quote
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 2022102 - Posted: 7 Dec 2019, 13:58:34 UTC - in response to Message 2022069.  

Are you saying that the powers-that-be undid the 100 max WUs?
I'm not - I keep my cache sizes low anyway, so it barely affects me. But that's what everyone else seems to be saying.

Even if true, we don't yet know whether the change was deliberate/permanent, or a mistake/temporary. Time will tell.


. . I suspect it is only temporary, but in the meantime setting work fetch to a moderate value seems the wise thing to me ...

Stephen

:)
ID: 2022102 · Report as offensive     Reply Quote
Profile Cliff Harding
Volunteer tester
Avatar

Send message
Joined: 18 Aug 99
Posts: 1432
Credit: 110,967,840
RAC: 67
United States
Message 2022105 - Posted: 7 Dec 2019, 14:14:15 UTC - in response to Message 2022102.  

I thought I was seeing things this morning, but it seems a lot of people are seeing the same thing. I did a quick calculation and it seems that the excess tasks are based on the number of working threads instead of the number of devices. My normal queue is 100/ CPU & 200/ GPU (1 CPU & 2 GPUs). I now have 198 CPU & 793 GPU tasks. That works out to approx. 2 x the number of working threads. Equals out to 4d of CPU & 6d of GPU work. Glad to have it & keep it coming..


I don't buy computers, I build them!!
ID: 2022105 · Report as offensive     Reply Quote
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 4267
Credit: 1,282,604,591
RAC: 6,640
United States
Message 2022113 - Posted: 7 Dec 2019, 14:46:29 UTC - in response to Message 2022101.  

Well, I'm down below 800 tasks on my 2 x GPU machine, and I'm back to 'no work available', with no mention of a limit. Looks like "Zalster's Theorem" is right, but I won't know for certain until I get back from lunch.


I think I concur with the 4x GPU limit now.

my "slow" system with [64] spoofed GPUs started going over the 6400 task limit. So I changed the system to [16] spoofed GPUs and now it's reporting that it has reached the limit for tasks in progress.
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 2022113 · Report as offensive     Reply Quote
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 2022114 - Posted: 7 Dec 2019, 14:49:36 UTC - in response to Message 2022113.  

I think I concur with the 4x GPU limit now.
my "slow" system with [64] spoofed GPUs started going over the 6400 task limit. So I changed the system to [16] spoofed GPUs and now it's reporting that it has reached the limit for tasks in progress.


. . But it is also raised for CPUs so someone else may be right that high thread CPUs can get more work too.

Stephen

. .
ID: 2022114 · Report as offensive     Reply Quote
Speedy
Volunteer tester
Avatar

Send message
Joined: 26 Jun 04
Posts: 1643
Credit: 12,921,799
RAC: 89
New Zealand
Message 2022263 - Posted: 8 Dec 2019, 0:25:43 UTC - in response to Message 2022258.  

My machine (Windows) managed to successfully get 79% more than normal cache size, normally 100 . I will return the _2 results 1st to help shrink the database
ID: 2022263 · Report as offensive     Reply Quote
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 35540
Credit: 261,360,520
RAC: 489
Australia
Message 2022312 - Posted: 8 Dec 2019, 4:45:20 UTC

With these new limits and my 3day+0.1 limit my 3570K rig gets a full cache of 1000 tasks while my 2500K rig (using -nobs while doing 1 less CPU task) sits at somewhere between 960-970 tasks.

Cheers.
ID: 2022312 · Report as offensive     Reply Quote
Previous · 1 · 2 · 3 · 4 · Next

Message boards : Number crunching : SETI@Home will now cache 150 CPU work units and 150 per installed GPU


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.