Panic Mode On (96) Server Problems?

Message boards : Number crunching : Panic Mode On (96) Server Problems?
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 17 · 18 · 19 · 20 · 21 · 22 · 23 . . . 24 · Next

AuthorMessage
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1661439 - Posted: 4 Apr 2015, 13:31:32 UTC - in response to Message 1661437.  

Yea mark I know, was a joke, just been so long since AP filled my cache. BUT woo hoo.
ID: 1661439 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5513
Credit: 528,817,460
RAC: 242
United States
Message 1661451 - Posted: 4 Apr 2015, 14:01:46 UTC - in response to Message 1661442.  
Last modified: 4 Apr 2015, 14:02:29 UTC

I can never fill my cache with APs. If anything, it seems as if my least capable machines are filled first, with my best only getting a few breadcrumbs :(

I still think they are handing out APs based on your ID number of your computer. I have yet to hear anyone says it isn't happening and explain why it isn't so..... :P

;) LOL
ID: 1661451 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1661454 - Posted: 4 Apr 2015, 14:09:24 UTC - in response to Message 1661451.  

As far as I know they only restrict sends more than X number if you are running a lot of errors. Which I'm sure your not because you watch the computers.

Other thing is your fast computers are processing them as fast as they can get them so, it's really hard to build a cache, but your slow computers can.
ID: 1661454 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1661459 - Posted: 4 Apr 2015, 14:14:52 UTC - in response to Message 1661451.  

But Zal I will trade you 150 AP's for a 980 to help with the whine *winks*
ID: 1661459 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5513
Credit: 528,817,460
RAC: 242
United States
Message 1661460 - Posted: 4 Apr 2015, 14:20:12 UTC - in response to Message 1661459.  

LOL,

Still after that 980 huh? ;) With all those people returning the 970s over the 3.5 GB memory issue, might be able to pick up a used one. That 3.5 hampers crunching Seti some but still is a good GPU for it even with that limitation.

No whine in a while since the backplates went on Ebay. I'm stuck at 60 APs in the cache and can't get more than that. My other machine with the 780s have 150 as well. Kind of weird.

It's going to be a mute point here soon. Looking at the server we are almost done with this run of APs then it will be back to MB for a while.

Zalster
ID: 1661460 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1661462 - Posted: 4 Apr 2015, 14:27:39 UTC - in response to Message 1661460.  
Last modified: 4 Apr 2015, 14:30:58 UTC

That brings up a question I have been meaning to ask here ... I have a 2GB card, but rarely see it using more than 750Mb (probably closer to 400-500) no matter what I throw at it, is that a command line thing to use more, or it will always rely on the CPU?
ID: 1661462 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5513
Credit: 528,817,460
RAC: 242
United States
Message 1661465 - Posted: 4 Apr 2015, 14:43:01 UTC - in response to Message 1661462.  

That is probable best to start a new thread for that topic.

But to answer, Work units use a combination of both physical and virtual RAM

The RAM on the card will determine how many work units you are have the card crunching at the same time (parallel, instances, 2up..whatever you want to call them) The more Ram the more each card can crunch at the same time.

Don't be fooled by seeing that it only uses x amount at 1 point in time. The amount can vary greatly during the course of the data being crunched.

When AP 7.04 came out, there were certain work units that were using over 1 GB by themselves. That stalled a lot of GPUs until they came out with v7.05

The last part of your question. Yes it will always use some of the CPU.

Think I explained that a long time ago somewhere as analogue of a factory

Think of the data a material on a conveyor belt. The more powerful the GPU the more belts it has with more workers looking at the data coming into the factory on the belts.

When it sees something that it thinks is interesting, it calls upon the foreman (CPU) and he comes over and examines that data and decided yes or no.

After that it either gets reported or moves along and the belt continues.

So a poor example but I think you get the idea.
ID: 1661465 · Report as offensive
Profile Brent Norman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2786
Credit: 685,657,289
RAC: 835
Canada
Message 1661472 - Posted: 4 Apr 2015, 14:59:14 UTC - in response to Message 1661465.  

Thanks for the explanation for GPU memory.

It was a typo on my part about using CPU, I meant RAM, I know the GPU needs the CPU to feed it. I have lots of RAM available, was thinking it was using some of it rather than GPU mem.
ID: 1661472 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 10914
Credit: 29,581,041
RAC: 66
United States
Message 1662013 - Posted: 6 Apr 2015, 16:41:23 UTC
Last modified: 6 Apr 2015, 16:41:54 UTC

Nobody has posted in this thread for over 2 days, is that a record?
ID: 1662013 · Report as offensive
Wild6-NJ
Volunteer tester

Send message
Joined: 4 Aug 99
Posts: 43
Credit: 100,336,791
RAC: 140
United States
Message 1662061 - Posted: 6 Apr 2015, 18:58:52 UTC - in response to Message 1662013.  

Nobody has posted in this thread for over 2 days, is that a record?


It's probably due to a sugar (aka AP) rush. ;)
ID: 1662061 · Report as offensive
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 24065
Credit: 261,360,520
RAC: 489
Australia
Message 1662405 - Posted: 8 Apr 2015, 0:54:47 UTC

Well the SSP hasn't updated in quite some time now and the cricket graph is almost on the deck plus my main rig's GPU's will be looking for other work soon. :-(

Cheers.
ID: 1662405 · Report as offensive
Profile cliff
Avatar

Send message
Joined: 16 Dec 07
Posts: 625
Credit: 3,590,440
RAC: 0
United Kingdom
Message 1662406 - Posted: 8 Apr 2015, 0:57:55 UTC - in response to Message 1662405.  

Hi Wiggo,

Well the SSP hasn't updated in quite some time now and the cricket graph is almost on the deck plus my main rig's GPU's will be looking for other work soon. :-(

Cheers.

Just tried to get some work, no go:-( Dispite server stats page showing loads of WU available:-(

Regards,
Cliff,
Been there, Done that, Still no damm T shirt!
ID: 1662406 · Report as offensive
Bill Butler
Avatar

Send message
Joined: 26 Aug 03
Posts: 101
Credit: 4,270,697
RAC: 0
United States
Message 1662409 - Posted: 8 Apr 2015, 1:12:13 UTC - in response to Message 1662405.  

Well the SSP hasn't updated in quite some time now . . .

I see the SSP update time is 22:10 UTC. Current time on the planet is 01:12. So, something is not cooperating.
"It is often darkest just before it turns completely black."
ID: 1662409 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5513
Credit: 528,817,460
RAC: 242
United States
Message 1662411 - Posted: 8 Apr 2015, 1:13:45 UTC - in response to Message 1662406.  

Had 1 batch of seti work right after the outage ended but didn't last long.

Since then been on the back up project.
ID: 1662411 · Report as offensive
Profile cliff
Avatar

Send message
Joined: 16 Dec 07
Posts: 625
Credit: 3,590,440
RAC: 0
United Kingdom
Message 1662418 - Posted: 8 Apr 2015, 1:35:25 UTC - in response to Message 1662411.  

Hi Zalster,

Had 1 batch of seti work right after the outage ended but didn't last long.

Since then been on the back up project.



Someone forgot to feed the hamsters:-)

Cheers,
Cliff,
Been there, Done that, Still no damm T shirt!
ID: 1662418 · Report as offensive
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 24065
Credit: 261,360,520
RAC: 489
Australia
Message 1662463 - Posted: 8 Apr 2015, 4:50:04 UTC

Other than a little bump in the CG it doesn't look like much has happened since my last post as the SSP is still the same. :-(

My main rig's GPU's will be requesting work from else where within 10mins.

Cheers.
ID: 1662463 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1662483 - Posted: 8 Apr 2015, 6:18:11 UTC

Watching that World Visualization thingy for the past 10 minutes, I saw 3 tasks sent out, as opposed to hundreds of results returned. All my boxes will make it to morning, but one will be empty by 1000 Berkeley time....
Donald
Infernal Optimist / Submariner, retired
ID: 1662483 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 1662490 - Posted: 8 Apr 2015, 6:55:25 UTC

There will be 44 new tasks as soon as the last task is finished here in a little while. I'm going to take this opportunity to clear out all the "lost" tasks, and there appear to be 44 on this host. Too bad they won't be sent back like back in the good old days...

I just noticed the last task sent has 21 Gaussians and then overflowed. I don't ever remember seeing that many Gaussians on one task before. Oh well, it's not like this task has never been run before...
http://setiathome.berkeley.edu/result.php?resultid=4081735035
ID: 1662490 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13270
Credit: 208,696,464
RAC: 304
Australia
Message 1662491 - Posted: 8 Apr 2015, 6:59:09 UTC - in response to Message 1662490.  

It appears the system is down.
No work available, and no updates on the server status- it's all still frozen.
Grant
Darwin NT
ID: 1662491 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13270
Credit: 208,696,464
RAC: 304
Australia
Message 1662500 - Posted: 8 Apr 2015, 7:51:47 UTC - in response to Message 1662491.  

Still borked.
Should be out of GPU work in a couple of hours.
Grant
Darwin NT
ID: 1662500 · Report as offensive
Previous · 1 . . . 17 · 18 · 19 · 20 · 21 · 22 · 23 . . . 24 · Next

Message boards : Number crunching : Panic Mode On (96) Server Problems?


 
©2022 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.