Panic Mode On (69) Server problems?

Message boards : Number crunching : Panic Mode On (69) Server problems?
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 6 · 7 · 8 · 9

AuthorMessage
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1201537 - Posted: 1 Mar 2012, 23:51:16 UTC - in response to Message 1201446.  

Hi Dave,
I'm betting thats not what Cosmic Ocean is thinking or feeling right now though..

Cheers

Eh.. it's not really that important. At first it was a project to compare the run-time vs. percent blanked for AP tasks using the lunatics apps. It started way back with r103 when I made the switch from stock->optimized. At that time, I had four machines that were crunching, and they were all radically different architectures. I made some good observations and data points.

Even recently when my main cruncher of just over five years started developing problems and I removed one of the CPUs, the data discovered a possible architecture flaw. I have at one point also sent all of my work to Josef to see if he could make any sense of an issue I was having.

So it wasn't really a waste, but like you said Dave.. it was probably time for that project to come to an end. I've worked through small periods of not being able to get at the tasks, or DB crashes that last a week or more without any significant loss, but this most recent occurrence was enough to just make me scrap it. Of course I could just start anew now that it is working for the most part.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1201537 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1202073 - Posted: 3 Mar 2012, 14:00:54 UTC - in response to Message 1202069.  

hey btw i have a question

i always asked myself , since i have the rescheduler, is it better to send all VHAR to CPU or the GPU ?

my GPU are making those in 1min 40sec but i m getting like 20-35 credits for those :P

would i get the same amount of credit if i keep them for my cores CPU ? (they are taking 33mins to do them)

On my machines (quad-core Intel CPUs, mostly elderly CUDA cards like 9800GT), I've always found the VHAR cunch particularly efficiently on the GPUs (ever since cuda23 came out, at least).

Running multiple VHAR on the CPUs tends to be counter-productive, because memory bus contention slows them all down.
ID: 1202073 · Report as offensive
Dave Stegner
Volunteer tester
Avatar

Send message
Joined: 20 Oct 04
Posts: 540
Credit: 65,583,328
RAC: 27
United States
Message 1202228 - Posted: 4 Mar 2012, 0:12:13 UTC

Anyone else getting this kind of stuff or is it just me ??

SLWS006

2953 SETI@home 3/3/2012 16:09:37 Sending scheduler request: Requested by user.
2954 SETI@home 3/3/2012 16:09:37 Reporting 2 completed tasks, not requesting new tasks
2955 3/3/2012 16:09:59 Project communication failed: attempting access to reference site
2956 SETI@home 3/3/2012 16:09:59 Scheduler request failed: Failure when receiving data from the peer
2957 3/3/2012 16:10:01 Internet access OK - project servers may be temporarily down.
2958 3/3/2012 16:10:02 Project communication failed: attempting access to reference site
2959 SETI@home 3/3/2012 16:10:02 Temporarily failed upload of 16dc11ad.8105.104986.7.10.124_1_0: HTTP error
2960 SETI@home 3/3/2012 16:10:02 Backing off 1 min 0 sec on upload of 16dc11ad.8105.104986.7.10.124_1_0
2961 3/3/2012 16:10:04 Internet access OK - project servers may be temporarily down.

Dave

ID: 1202228 · Report as offensive
Profile Dimly Lit Lightbulb 😀
Volunteer tester
Avatar

Send message
Joined: 30 Aug 08
Posts: 15399
Credit: 7,423,413
RAC: 1
United Kingdom
Message 1202243 - Posted: 4 Mar 2012, 1:35:34 UTC - in response to Message 1202232.  

I think you MB people are about to run out of new MB WU's. Not many files left to split now. Only 35 channels left to do.

There will be whining soon, I'm sure :-)

As an astropulse only cruncher, I shall giggle that the multibeamers will have the experience of what it's like for no tasks to be split for them to crunch :)

Member of the People Encouraging Niceness In Society club.

ID: 1202243 · Report as offensive
Profile Dimly Lit Lightbulb 😀
Volunteer tester
Avatar

Send message
Joined: 30 Aug 08
Posts: 15399
Credit: 7,423,413
RAC: 1
United Kingdom
Message 1202251 - Posted: 4 Mar 2012, 2:22:36 UTC - in response to Message 1202244.  

I think you MB people are about to run out of new MB WU's. Not many files left to split now. Only 35 channels left to do.

There will be whining soon, I'm sure :-)

As an astropulse only cruncher, I shall giggle that the multibeamers will have the experience of what it's like for no tasks to be split for them to crunch :)


There is a high chance that I will join you in that giggle :-)

Should we put it down as Seti Karma? :)

Member of the People Encouraging Niceness In Society club.

ID: 1202251 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1202268 - Posted: 4 Mar 2012, 3:46:04 UTC - in response to Message 1202228.  

Anyone else getting this kind of stuff or is it just me ??

Maybe just you.
I've hit the server limits, but when i can get work it comes down at a good pace.

Grant
Darwin NT
ID: 1202268 · Report as offensive
Profile Belthazor
Volunteer tester
Avatar

Send message
Joined: 6 Apr 00
Posts: 219
Credit: 10,373,795
RAC: 13
Russia
Message 1202289 - Posted: 4 Mar 2012, 6:01:11 UTC - in response to Message 1202244.  


As an astropulse only cruncher, I shall giggle that the multibeamers will have the experience of what it's like for no tasks to be split for them to crunch :)


There is a high chance that I will join you in that giggle :-)


I believe every cruncher have a nice bunch now :-P
ID: 1202289 · Report as offensive
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 34744
Credit: 261,360,520
RAC: 489
Australia
Message 1202293 - Posted: 4 Mar 2012, 6:51:35 UTC - in response to Message 1202289.  

Seeing as even with the limits imposed I'm still good for 3 days so I can't see me whining any time soon. :D

My "whining" will likely start when the AP's are turned back on and I have to go searching for good proxies again. ;)

Cheers.
ID: 1202293 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1202294 - Posted: 4 Mar 2012, 7:05:33 UTC - in response to Message 1202293.  

Seeing as even with the limits imposed I'm still good for 3 days

I wish.
3-3.5 days CPU. Only 0.8-1.1 for the GPUs. I can't be bothered messing around with the FLOPS numbers to get the DCF to stabilise.
Grant
Darwin NT
ID: 1202294 · Report as offensive
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 34744
Credit: 261,360,520
RAC: 489
Australia
Message 1202297 - Posted: 4 Mar 2012, 7:13:12 UTC - in response to Message 1202294.  

Seeing as even with the limits imposed I'm still good for 3 days

I wish.
3-3.5 days CPU. Only 0.8-1.1 for the GPUs. I can't be bothered messing around with the FLOPS numbers to get the DCF to stabilise.

I have never had the inclination or time to bother with flops either Grant, I just let mine sort themselves out (hopefully Dr D.A. will start to show that he's worth his wages doing the right thing instead of just changing things to make it look like he's worth it).

Cheers.
ID: 1202297 · Report as offensive
Profile Belthazor
Volunteer tester
Avatar

Send message
Joined: 6 Apr 00
Posts: 219
Credit: 10,373,795
RAC: 13
Russia
Message 1202310 - Posted: 4 Mar 2012, 8:26:24 UTC

BTW, one tape was added just now. 35 channels to split. No items to whine at all - multibeam forever! ;-)))
ID: 1202310 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1202342 - Posted: 4 Mar 2012, 13:40:24 UTC - in response to Message 1202310.  
Last modified: 4 Mar 2012, 13:41:54 UTC

BTW, one tape was added just now. 35 channels to split. No items to whine at all - multibeam forever! ;-)))

At present (2012-03-04 13:30:07 UTC) the counts are at. total channels to do: MB 54/84 AP 98/98

So until the AP start again I suspect the whining will be held down to a minimum.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1202342 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1202409 - Posted: 4 Mar 2012, 18:04:44 UTC - in response to Message 1202407.  

It's going to be interesting to see how close to zero results out in the field we get for AP, before they start sending out new AP workunits. As of now, we have 86,770 AP results out in the field.

It could be that the next time the AP splitters run here, they could be splitting AP v6 Wu's instead of v505 Wu's, we'll see in time,

Claggy
ID: 1202409 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1202412 - Posted: 4 Mar 2012, 18:14:05 UTC - in response to Message 1202410.  

It's going to be interesting to see how close to zero results out in the field we get for AP, before they start sending out new AP workunits. As of now, we have 86,770 AP results out in the field.

It could be that the next time the AP splitters run here, they could be splitting AP v6 Wu's instead of v505 Wu's, we'll see in time,

Claggy


Wouldn't that require new apps for both the CPU, and for the ATI OpenCL? In which case I ask:

Where's my new apps for AP v6?

LOL

You'll get them when you need them ;-)

Claggy
ID: 1202412 · Report as offensive
Previous · 1 . . . 6 · 7 · 8 · 9

Message boards : Number crunching : Panic Mode On (69) Server problems?


 
©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.