Advanced search

Message boards : Server and website : Are we out of long runs?

Author Message
flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,550,610,997
RAC: 67,223
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 32298 - Posted: 26 Aug 2013 | 17:45:19 UTC

I have 1 machine sitting idle, my log says "No work to send".

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2078
Credit: 15,144,101,390
RAC: 4,393,617
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32300 - Posted: 26 Aug 2013 | 18:20:44 UTC - in response to Message 32298.

I have 1 machine sitting idle, my log says "No work to send".

The short (and the beta) queue is empty at the moment.

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,550,610,997
RAC: 67,223
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 32311 - Posted: 27 Aug 2013 | 1:06:10 UTC

It would appear I was right, any idea when we'll be getting more work? Did those cancellations also effect wu's waiting to be sent? There were over 2000 just a couple of days ago.

werdwerdus
Send message
Joined: 15 Apr 10
Posts: 123
Credit: 1,004,473,861
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32312 - Posted: 27 Aug 2013 | 4:25:14 UTC

drew-PC

11637 GPUGRID 8/26/2013 11:26:22 PM No tasks are available for Short runs (2-3 hours on fastest card)
11638 GPUGRID 8/26/2013 11:26:22 PM No tasks are available for Long runs (8-12 hours on fastest card)

____________
XtremeSystems.org - #1 Team in GPUGrid

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1919
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 32314 - Posted: 27 Aug 2013 | 5:20:22 UTC - in response to Message 32312.

Coming soon.

gdf

Stefan
Volunteer moderator
Project developer
Project scientist
Send message
Joined: 5 Mar 13
Posts: 348
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 32316 - Posted: 27 Aug 2013 | 8:25:56 UTC

I am also sending right now another 300 of my Adapt series.

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,550,610,997
RAC: 67,223
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 32346 - Posted: 28 Aug 2013 | 4:30:05 UTC

Am I the only one who notices when were out or does everybody just sit around waiting for me to say something? It's OK, I'll let someone else take over for a while.

Profile dskagcommunity
Avatar
Send message
Joined: 28 Apr 11
Posts: 456
Credit: 815,576,358
RAC: 0
Level
Glu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32348 - Posted: 28 Aug 2013 | 5:53:44 UTC
Last modified: 28 Aug 2013 | 5:57:02 UTC

I dont post outtages anymore because admin team said they have machanisem to see when work getting short. So i trust on that ;) But i got sometimes no work too, or i get sometimes emergency short units (so im under 800k RAC). Your not alone ^^
____________
DSKAG Austria Research Team: http://www.research.dskag.at



Crunching for my deceased Dog who had "good" Braincancer..

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,550,610,997
RAC: 67,223
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 32349 - Posted: 28 Aug 2013 | 7:05:04 UTC

Well, I know the page that shows how much is left in the queue takes a snap shot every (you know this though) hour or so. If it's empty, it (the web page) could show there are still wu's ready to go. I don't know, I guess I'll stop posting about it, I figured there might be somebody around there. I'm not upset at all, just pointing it out.

Profile dskagcommunity
Avatar
Send message
Joined: 28 Apr 11
Posts: 456
Credit: 815,576,358
RAC: 0
Level
Glu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32350 - Posted: 28 Aug 2013 | 7:27:00 UTC

I would love more to see boinc manager asking more often for tasks...after hitting 4 times the refresh button i get work on a idle machine..why cant do the manager that for me -_-
____________
DSKAG Austria Research Team: http://www.research.dskag.at



Crunching for my deceased Dog who had "good" Braincancer..

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,550,610,997
RAC: 67,223
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 32351 - Posted: 28 Aug 2013 | 7:35:56 UTC

Maybe Zoltan could write a script for you, I'll bet he knows how to do it. I'm just glad there not crashing any more

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2078
Credit: 15,144,101,390
RAC: 4,393,617
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32355 - Posted: 28 Aug 2013 | 8:49:12 UTC - in response to Message 32350.

I would love more to see boinc manager asking more often for tasks...after hitting 4 times the refresh button i get work on a idle machine..why cant do the manager that for me -_-

The manager has to increase the period of every connection attempt when there's no work on the server, or else the clients asking for new work would make so many request to the server like a DDOS attack, when all the clients run out of work.

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,550,610,997
RAC: 67,223
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 32359 - Posted: 28 Aug 2013 | 9:54:08 UTC - in response to Message 32355.
Last modified: 28 Aug 2013 | 9:55:08 UTC

Ha, that makes sense, I upgraded my other 3 rigs to the 326.84 driver, might have to add the short queue.

Profile dskagcommunity
Avatar
Send message
Joined: 28 Apr 11
Posts: 456
Credit: 815,576,358
RAC: 0
Level
Glu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32360 - Posted: 28 Aug 2013 | 10:12:28 UTC

I dont think so, when there are worrys about ddos then they only need to set a as example 5 minute limit against the 30 secs from what its set at the moment. Then it is unpossible that there are too much requests at the same time. Or the boinc manager should write down into his window that he increased the time. So it often looks like he dont want to get work from the primary projects. But thats a boinc general thing nothing about gpugrid, they have often enough a steady amount of work :) was only wondering if there is a trick or something for unattended machines ^^
____________
DSKAG Austria Research Team: http://www.research.dskag.at



Crunching for my deceased Dog who had "good" Braincancer..

ATF
Avatar
Send message
Joined: 5 Jan 12
Posts: 117
Credit: 40,721,760
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwat
Message 32372 - Posted: 28 Aug 2013 | 12:25:59 UTC

I didnt read much, but i did start the energy project on world community grid as im not using this computer much. The thing is when I do want to use it I have to suspend and resume manually. You cant say run the gpu nonstop and not run the cpu non stop. Kinda backwards if you ask me. But hey they also make a 5ghz processor that takes up 300 watts.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2078
Credit: 15,144,101,390
RAC: 4,393,617
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32378 - Posted: 28 Aug 2013 | 13:58:44 UTC

It's very odd, that I'm receiving:
2013. 08. 28. 15:45:24 GPUGRID Message from server: No tasks sent
2013. 08. 28. 15:45:24 GPUGRID Message from server: No tasks are available for ACEMD beta version
2013. 08. 28. 15:45:24 GPUGRID Message from server: No tasks are available for Long runs (8-12 hours on fastest card)
While the server status page shows 527 unsent long workunits at the moment.

Stefan
Volunteer moderator
Project developer
Project scientist
Send message
Joined: 5 Mar 13
Posts: 348
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 32379 - Posted: 28 Aug 2013 | 14:17:40 UTC - in response to Message 32378.
Last modified: 28 Aug 2013 | 14:30:19 UTC

It *might* have to do with WU priorities. I had to send my WU's with very high priority because we need the results extremely soon. However from what I understand, when the priority is very high, WU's get automatically sent only to users with very fast turnaround I guess to avoid sending them to slower machines which might keep them "delayed" for a longer time.

That is my guess on why you don't get any.

HA-SOFT, s.r.o.
Send message
Joined: 3 Oct 11
Posts: 100
Credit: 5,536,348,501
RAC: 67,394
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32380 - Posted: 28 Aug 2013 | 14:24:48 UTC - in response to Message 32379.

My PC has turnaround time 0.27 (680) and 0.21 (Titan) and not getting any tasks.

What is faster turnaround than this?

Stefan
Volunteer moderator
Project developer
Project scientist
Send message
Joined: 5 Mar 13
Posts: 348
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 32381 - Posted: 28 Aug 2013 | 14:26:56 UTC - in response to Message 32380.
Last modified: 28 Aug 2013 | 14:27:55 UTC

As I said that was only my theory. I didn't mean to understate your turnaround times which are probably great :D I have no practical experience with the BOINC scheduler. Maybe someone with more experience can answer the question better.

HA-SOFT, s.r.o.
Send message
Joined: 3 Oct 11
Posts: 100
Credit: 5,536,348,501
RAC: 67,394
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32382 - Posted: 28 Aug 2013 | 14:30:53 UTC - in response to Message 32381.

BOINC scheduler is some kind of mysterious thing. It lives on its own.

flashawk
Send message
Joined: 18 Jun 12
Posts: 297
Credit: 3,550,610,997
RAC: 67,223
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 32385 - Posted: 28 Aug 2013 | 15:23:27 UTC

I don't think the server is giving out work units, the number of LR's has gone up and I have 2 cards sitting idol.

TJ
Send message
Joined: 26 Jun 09
Posts: 815
Credit: 1,470,348,955
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32387 - Posted: 28 Aug 2013 | 15:31:54 UTC - in response to Message 32385.

I don't think the server is giving out work units, the number of LR's has gone up and I have 2 cards sitting idol.

Yeah indeed we need to wait until a WU reports back errorred and then the first rig that request new work at exact the right time get it to try over.
____________
Greetings from TJ

Stefan
Volunteer moderator
Project developer
Project scientist
Send message
Joined: 5 Mar 13
Posts: 348
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 32389 - Posted: 28 Aug 2013 | 15:49:23 UTC

Something is fishy. None of the WU's I sent today are being processed and from what I see the processed ones are going down while the unsent are going up.
We are looking into it.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2078
Credit: 15,144,101,390
RAC: 4,393,617
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32390 - Posted: 28 Aug 2013 | 15:52:02 UTC

Since I've opted out beta tasks, I haven't received work (long runs)....

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2078
Credit: 15,144,101,390
RAC: 4,393,617
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32393 - Posted: 28 Aug 2013 | 16:02:55 UTC
Last modified: 28 Aug 2013 | 16:03:17 UTC

2013. 08. 28. 18:02:50 GPUGRID Message from server: Project is temporarily shut down for maintenance
Guys, be careful with that :)

Carlos Augusto Engel
Send message
Joined: 5 Jun 09
Posts: 38
Credit: 2,668,296,104
RAC: 335,298
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32394 - Posted: 28 Aug 2013 | 16:03:27 UTC - in response to Message 32390.

Since I've opted out beta tasks, I haven't received work (long runs)....

But i did:
7217066 4727967 28 Aug 2013 | 7:23:43 UTC 2 Sep 2013 | 7:23:43 UTC Em progresso --- --- --- ACEMD beta version v8.00 (cuda55)
7216368 4726462 27 Aug 2013 | 22:56:33 UTC 1 Sep 2013 | 22:56:33 UTC Em progresso --- --- --- Long runs (8-12 hours on fastest card) v8.00 (cuda55)
7216367 4727076 27 Aug 2013 | 22:55:54 UTC 28 Aug 2013 | 11:15:45 UTC Completo e validado 6,418.41 6,008.85 60,900.00 ACEMD beta version v8.00 (cuda55)

____________

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2078
Credit: 15,144,101,390
RAC: 4,393,617
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32395 - Posted: 28 Aug 2013 | 16:04:41 UTC

2013. 08. 28. 18:05:05 GPUGRID Requesting new tasks for CPU and GPU
2013. 08. 28. 18:05:06 GPUGRID Scheduler request completed: got 1 new tasks
2013. 08. 28. 18:05:08 GPUGRID Started download of e7s2_e5s5f56-SDOERR_BenAdaptNN9-0-LICENSE
Success! :)

HA-SOFT, s.r.o.
Send message
Joined: 3 Oct 11
Posts: 100
Credit: 5,536,348,501
RAC: 67,394
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32397 - Posted: 28 Aug 2013 | 16:06:27 UTC - in response to Message 32394.

Me too.

Stefan
Volunteer moderator
Project developer
Project scientist
Send message
Joined: 5 Mar 13
Posts: 348
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 32398 - Posted: 28 Aug 2013 | 16:08:32 UTC
Last modified: 28 Aug 2013 | 16:08:44 UTC

We fixed it. It was actually a combination of problems. My WU's had high priority and the credentials of the users had somehow been lost. So noone was fast enough for my WU's haha :D
It's fixed now though.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2078
Credit: 15,144,101,390
RAC: 4,393,617
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32399 - Posted: 28 Aug 2013 | 16:11:11 UTC

Thank you.

Profile dskagcommunity
Avatar
Send message
Joined: 28 Apr 11
Posts: 456
Credit: 815,576,358
RAC: 0
Level
Glu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32401 - Posted: 28 Aug 2013 | 16:31:28 UTC

Great news! no more hacking the refresh button :D
____________
DSKAG Austria Research Team: http://www.research.dskag.at



Crunching for my deceased Dog who had "good" Braincancer..

Profile Stoneageman
Avatar
Send message
Joined: 25 May 09
Posts: 213
Credit: 15,703,683,088
RAC: 1,393,188
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32577 - Posted: 1 Sep 2013 | 1:16:29 UTC
Last modified: 1 Sep 2013 | 1:20:08 UTC

Bone dry again on all queues

nanoprobe
Send message
Joined: 26 Feb 12
Posts: 181
Credit: 222,317,747
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 32578 - Posted: 1 Sep 2013 | 5:53:41 UTC - in response to Message 32577.
Last modified: 1 Sep 2013 | 6:07:21 UTC

Bone dry again on all queues

You finished 'em all SAM.

Profile dskagcommunity
Avatar
Send message
Joined: 28 Apr 11
Posts: 456
Credit: 815,576,358
RAC: 0
Level
Glu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32582 - Posted: 1 Sep 2013 | 9:09:47 UTC
Last modified: 1 Sep 2013 | 9:12:36 UTC

No wonder titans and 780s dry up all :p or was the 1800workunits in work a normal value? I think it was much lower or? :) thats always the positive thing with slower machines on shortages, the system has longer to do with a limited cache like gpugrid and poem ^^
____________
DSKAG Austria Research Team: http://www.research.dskag.at



Crunching for my deceased Dog who had "good" Braincancer..

TJ
Send message
Joined: 26 Jun 09
Posts: 815
Credit: 1,470,348,955
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 32597 - Posted: 1 Sep 2013 | 16:12:48 UTC - in response to Message 32582.

No wonder titans and 780s dry up all :p or was the 1800workunits in work a normal value? I think it was much lower or? :) thats always the positive thing with slower machines on shortages, the system has longer to do with a limited cache like gpugrid and poem ^^

One day we will replace our GTX285 with Titans ultra dskagcommunity and then we have super-rigs too ;-)
____________
Greetings from TJ

Post to thread

Message boards : Server and website : Are we out of long runs?