Advanced search

Message boards : Graphics cards (GPUs) : No work issues - update 31Dec 08

Author Message
Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 5111 - Posted: 31 Dec 2008 | 13:06:05 UTC


So, the situation is that since mid December the BOINC server has been changed with a new scheduling policy.

Before that everything was working pretty smoothly. Probably, the necessity to handle more complicated situations like two projects pushed for the change.

Unfortunately, this new server policy seems to work less than optimally and hosts remain without work. Several ways to try to bypass the problem are available in our forum. But for now until BOINC developers find a solution, we have to manage this way.

We have asked if it is worth to back-port the server to a version before mid December. Most likely, we will wait until next week after the holiday season when BOINC dev should be able to address the problems once for all.

I hope, as much as you, that this is fixed soon, as it is creating more maintenance work and reduced throughput for the project.

For now, happy new year to everybody.
GPUGRID team.

Profile K1atOdessa
Send message
Joined: 25 Feb 08
Posts: 249
Credit: 370,320,941
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5112 - Posted: 31 Dec 2008 | 13:20:22 UTC - in response to Message 5111.

Thanks, GDF, for the update.

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5127 - Posted: 31 Dec 2008 | 16:52:12 UTC

Yes ... Thank you!

I have been wondering if we are part of an expiriment with random rewards.

I was close to running out on my main box and so I started to do all the things that suppsidly help ... no luck ... then I just tried again randomly and got three ...

So, I am starting to feel like my dog Pixie, who knows that if she just scratches at the door it will open ... of course if she would just wait for the time it would open for her anyway ... but, she KNOWS that scratching at that automatic door will make it work ...

Sigh ...

Anyway, I have work until tomorrow ...

Profile [BOINC@Poland]AiDec
Send message
Joined: 2 Sep 08
Posts: 53
Credit: 9,213,937
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwat
Message 5136 - Posted: 1 Jan 2009 | 13:40:48 UTC
Last modified: 1 Jan 2009 | 13:42:08 UTC

My humble opinion - don`t go roll back soft on the server. W8 until problem will be solved.

We`ve already found some possibilities how to manage work for our maschines. We`ve learned how to live in new world :P. Double OS (any) on one maschine + VM with next three OS`s - each one can store 4 WU`s in stock, and e.g. 300 WU`s SETIbeta in stock with resource share 1 compare to 500 in GPUGrid (means SETI will be crunched just if there is no WU`s from GPUGrid).


And now I believe we can w8 :).
____________

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5137 - Posted: 1 Jan 2009 | 14:56:36 UTC

I got the no work message a couple times.

Then with only 30 min to go before I would run dry I got 2 automatically.

Then, later on when reporting a task I got another ... just as if it was working ...

Now I have one in work and 2 pending ... enough to last until this evening I think ...

Well, time marches on ...

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5139 - Posted: 1 Jan 2009 | 15:10:15 UTC
Last modified: 1 Jan 2009 | 15:10:45 UTC

I think this thread is perfect to thank the GPU-Grid team. Thanks for giving us this interesting project, for the ability to use the power of GPUs under BOINC and thanks for the continued support even during the holidays! And all the best for the new year :)

MrS
____________
Scanning for our furry friends since Jan 2002

Profile K1atOdessa
Send message
Joined: 25 Feb 08
Posts: 249
Credit: 370,320,941
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5140 - Posted: 1 Jan 2009 | 15:44:16 UTC - in response to Message 5139.

I think this thread is perfect to thank the GPU-Grid team. Thanks for giving us this interesting project, for the ability to use the power of GPUs under BOINC and thanks for the continued support even during the holidays! And all the best for the new year :)

MrS


I second that.

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1112
Credit: 6,162,416,256
RAC: 0
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5141 - Posted: 1 Jan 2009 | 15:50:19 UTC - in response to Message 5140.

I think this thread is perfect to thank the GPU-Grid team. Thanks for giving us this interesting project, for the ability to use the power of GPUs under BOINC and thanks for the continued support even during the holidays! And all the best for the new year :)

I second that.

Yes! Thanks for the GREAT project!

Profile [XTBA>XTC] D.I.Y.Calculat...
Send message
Joined: 2 Sep 08
Posts: 4
Credit: 15,705,436
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwat
Message 5154 - Posted: 1 Jan 2009 | 17:27:58 UTC - in response to Message 5141.

Yes, thank's to this team

Milford
Send message
Joined: 17 Jul 07
Posts: 14
Credit: 9,618,510
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwatwatwatwat
Message 5155 - Posted: 1 Jan 2009 | 17:38:02 UTC

And a Happy New Year!
____________

Profile [EK]Maniac
Send message
Joined: 14 Jun 08
Posts: 6
Credit: 4,448,982
RAC: 0
Level
Ala
Scientific publications
watwatwatwat
Message 5156 - Posted: 1 Jan 2009 | 18:14:49 UTC - in response to Message 5155.

Happy new year to GPUGrid and all it's user

Many thnx for the support, everything works fine

Maniac
____________

Profile Bender10
Avatar
Send message
Joined: 3 Dec 07
Posts: 167
Credit: 8,368,897
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwatwat
Message 5308 - Posted: 5 Jan 2009 | 14:55:18 UTC

My system has completed (and recieved) 2 wu's, without any manual intervention from me.

Server problem fixed..?

Linux64, 6.3.21, 177.xx driver
____________


Consciousness: That annoying time between naps......

Experience is a wonderful thing: it enables you to recognize a mistake every time you repeat it.

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5309 - Posted: 5 Jan 2009 | 17:10:09 UTC - in response to Message 5308.
Last modified: 5 Jan 2009 | 18:05:41 UTC

My system has completed (and recieved) 2 wu's, without any manual intervention from me.

Server problem fixed..?

Linux64, 6.3.21, 177.xx driver


I don't think so ... I have had random successes in the past also ... it has never lasted...

Thats the whole point to the random pigeon experiments ... randomly reward and rituals develop ... I have 4 more in queue and one in work and if it was all working I should have about 6 in queue ...

The good news is that I have 80 days of CEP awarded and a little over 9 days pending and a couple finishing up ... though I have no idea how long I will have to wait for the wingmen to come in ... I also probably need to build up a couple spares just in case ... :)

But I have drained SIMAP and they are out of work so on to Cosmology and ABC...


{edit}Maybe I was wrong... it looks like I did an autofetch of a new task ... now if it keeps up for a couple days .... {/edit}

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 5340 - Posted: 6 Jan 2009 | 13:20:56 UTC - in response to Message 5309.

So, thanks for all the messages.

Now, it seems that there was a problem in the configuration of the server, so that the queue was filled up with jobs for the PS3. We have now changed to distribute WUs equally among applications.

Please check that this is the case, so that we can announce that the problem is fixed.

gdf.

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5342 - Posted: 6 Jan 2009 | 13:43:35 UTC - in response to Message 5340.

So, thanks for all the messages.

Now, it seems that there was a problem in the configuration of the server, so that the queue was filled up with jobs for the PS3. We have now changed to distribute WUs equally among applications.

Please check that this is the case, so that we can announce that the problem is fixed.

gdf.


I hate being right all the time ...

The last two days I have been getting tasks on demand and my queues are stable with the appropriate number of tasks for each system. When my systems request work it seems to get work as it should to keep the customer satisfied.

So, I think we may have gotten it ... the problem is that I have had these streaks before ... :)

But they have never gone on so long ... another day and if my i7 still has 5 tasks pending as it should, and I have done no manual interventions, well, I say Bravo-Zulu ... Well Done!

Scott Brown
Send message
Joined: 21 Oct 08
Posts: 144
Credit: 2,973,555
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwatwat
Message 5343 - Posted: 6 Jan 2009 | 13:53:49 UTC


I am also seeing good work flow with both a low-end card and a mid-range card for the last couple of days.

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 1,088
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 5345 - Posted: 6 Jan 2009 | 14:42:43 UTC

It seems to work pretty nice for me for the last two days. All three GPU capable PCs have a full queue of GPUGRID tasks now.

Only yesterdy evening my Vista PC ran dry of workunits and I got the "no Cell work for you" message again, paired with a 24h deferral. A manual update a half hour later fixed that and I got two tasks again...

I'll watch my computers closely if it happens again.
____________

pixelicious.at - my little photoblog

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5362 - Posted: 7 Jan 2009 | 11:18:46 UTC

Still working ... I had two tasks ready to report ... did update and back to 4 pending and one in work (I lowered my work queue size) ... do, it seems to have been fixed.

Now, if we could only have the BOINC Manager problems fixed I would be really happy ...

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 5364 - Posted: 7 Jan 2009 | 13:41:45 UTC

Seems to be working better the last few days for me too, all my Caches are full at least & I only have to manually Update once in awhile to fill up a Cache ... :)

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5366 - Posted: 7 Jan 2009 | 15:24:06 UTC - in response to Message 5364.

Seems to be working better the last few days for me too, all my Caches are full at least & I only have to manually Update once in awhile to fill up a Cache ... :)


Almost makes one want to go out and buy some faster cards ... :)

Sadly, I have to pay for the i7 first ... ah well ... that bill will be here soon enough ...

Profile K1atOdessa
Send message
Joined: 25 Feb 08
Posts: 249
Credit: 370,320,941
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5367 - Posted: 7 Jan 2009 | 15:33:16 UTC - in response to Message 5364.

Seems to be working better the last few days for me too, all my Caches are full at least & I only have to manually Update once in awhile to fill up a Cache ... :)


Same here. The only reason I had to manually update recently is the "not available for your type of processor" message I've received occasionally. I think this is due to the low # of WU's (< 100 last night) waiting to be sent out and not necessarily the PS3 / GPU queue issue.

Other than that, after the client sends back a result it usually obtains a new WU in short order.

localizer
Send message
Joined: 17 Apr 08
Posts: 113
Credit: 1,656,514,857
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5368 - Posted: 7 Jan 2009 | 18:01:13 UTC

............ better but not great. I still have to suspend all other projects to avoid the 'no work for your type of processor message' and the 24 hour 'cooling off' period is still not appreciated. Why do we have to have this long a period?.
Had a couple of 'computation errors' in the last 24 hours - could be local, but is also new.

Two steps forward, one step back.

P.

Profile [FVG] bax
Avatar
Send message
Joined: 18 Jun 08
Posts: 29
Credit: 17,772,874
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwat
Message 5369 - Posted: 7 Jan 2009 | 18:32:00 UTC

It's working properly for me

2 PS3 and 2 GPUs (GTX) with full load

Profile Bender10
Avatar
Send message
Joined: 3 Dec 07
Posts: 167
Credit: 8,368,897
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwatwat
Message 5402 - Posted: 9 Jan 2009 | 0:21:33 UTC - in response to Message 5369.

Has anyone noticed their Wu cache being updated without user intervention..? Mine seems to doing ok for the last 24-36 hours...
____________


Consciousness: That annoying time between naps......

Experience is a wonderful thing: it enables you to recognize a mistake every time you repeat it.

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5409 - Posted: 9 Jan 2009 | 4:32:02 UTC - in response to Message 5402.

Has anyone noticed their Wu cache being updated without user intervention..? Mine seems to doing ok for the last 24-36 hours...



Yes ...

Actually, it has been 36-48 hours ... we commented about this in other threads ... as in ... ITS FIXED!


Or something like that ... :)

Profile mike047
Send message
Joined: 21 Dec 08
Posts: 47
Credit: 7,330,049
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwatwat
Message 5413 - Posted: 9 Jan 2009 | 8:09:01 UTC - in response to Message 5402.

Has anyone noticed their Wu cache being updated without user intervention..? Mine seems to doing ok for the last 24-36 hours...


My boxes have worked without help for several days now:D

mike

Profile Kokomiko
Avatar
Send message
Joined: 18 Jul 08
Posts: 190
Credit: 24,093,690
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 5425 - Posted: 9 Jan 2009 | 19:12:10 UTC

Same here, no problems since the error on the server is fixed. For those, which has still problems: Give GPUGrid a little bit more calculation time then the percentage of the GPU feeding core in relation to the other cores.

Example:
1 Quad core with 1 GPU at 4+1 -> GPUGrid needs more than 25% to run without problems.

1 DualCore with 1 GPU at 2+1 -> GPUGrid needs more than 50% to run without problems.

1 Quad core with 2 GPUs at 4+2 -> give over 50% to GPUGrid.
____________

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 5443 - Posted: 10 Jan 2009 | 13:23:35 UTC

Same here, no problems since a few days. And since I switched from 6.4.2 to 6.5.0 there's no babysitting either, BOINC just fetches new work according to the cache setting *being a little amazed*.

MrS
____________
Scanning for our furry friends since Jan 2002

Post to thread

Message boards : Graphics cards (GPUs) : No work issues - update 31Dec 08

//