Panic Mode On (13) Server problems

Message boards : Number crunching : Panic Mode On (13) Server problems
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 14 · Next

AuthorMessage
Alinator
Volunteer tester

Send message
Joined: 19 Apr 05
Posts: 4178
Credit: 4,647,982
RAC: 0
United States
Message 868212 - Posted: 22 Feb 2009, 21:00:18 UTC - in response to Message 868195.  
Last modified: 22 Feb 2009, 21:01:20 UTC

Well, you're right that there will be a pantload of tasks to upload when the jam finally breaks. Fortunately, the actual file to be sent is fairly small so it should clear relatively fast (typically it takes far less than a day).

Download backlogs are the real killers, since they can range from a couple of hundred K to megabytes depending on what's waiting.

Alinator
ID: 868212 · Report as offensive
nemesis
Avatar

Send message
Joined: 12 Oct 99
Posts: 1408
Credit: 35,074,350
RAC: 0
Message 868215 - Posted: 22 Feb 2009, 21:03:51 UTC

definately MB's this time.....
ID: 868215 · Report as offensive
OzzFan Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Apr 02
Posts: 15691
Credit: 84,761,841
RAC: 28
United States
Message 868225 - Posted: 22 Feb 2009, 21:26:04 UTC - in response to Message 868179.  

Might have to switch to Milky Way. This is the first project that i encountered upload problems.


I'm guessing you don't participate in may projects then. I've had upload problems with Rosetta too.

This would be exactly why SETI recommends, per the front page, that you join multiple projects. They never said things would be smooth sailing at all times.
ID: 868225 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 868228 - Posted: 22 Feb 2009, 21:29:31 UTC - in response to Message 868194.  
Last modified: 22 Feb 2009, 21:30:57 UTC

Have not been able to upload for about 16 hours now, more than 500 WU to upload. 4 day cache, so i guess it is not an issue at the moment.

BUT:

Looking at the server status page and Scarecrow graphs makes me wonder.

Normally MB WUs return at the rate of 10 to 20 to 1 AP WU.

Why is it suddenly appear easier to return an AP work unit than an MB WU ??

Presently the ratio is 2 to 1, prox.


Bandwidth is maxed out at Seti@Home, I terminated My network activity in Boinc until much later. To Increase bandwidth I've read would cost $80,000.00 total.

Go here and read and maybe donate to the cause and If one does donate, Make It a specific donation.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 868228 · 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 868248 - Posted: 22 Feb 2009, 22:03:05 UTC

I realize that bandwidth is maxed, that was not my point.


Looking at the server status page and Scarecrow graphs makes me wonder.

Normally MB WUs return at the rate of 10 to 20 to 1 AP WU.

Why does it suddenly appear easier to return an AP work unit than an MB WU ??

Presently the ratio is 2 to 1, prox.


Dave

ID: 868248 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 868255 - Posted: 22 Feb 2009, 22:18:15 UTC - in response to Message 868195.  
Last modified: 22 Feb 2009, 22:21:12 UTC

What sort of timescale are we looking at to clear the backlog.

If the problem were to resolve itself now, i'd say about 8 hours.
Grant
Darwin NT
ID: 868255 · Report as offensive
Tribble

Send message
Joined: 21 Feb 02
Posts: 65
Credit: 7,978,002
RAC: 0
Australia
Message 868261 - Posted: 22 Feb 2009, 22:24:46 UTC - in response to Message 868255.  

My CPU core temp has dropped to 20 degrees with nothing to do.
How am I meant to save heating costs with no WUs? :P
ID: 868261 · Report as offensive
Profile Ghery S. Pettit
Avatar

Send message
Joined: 7 Nov 99
Posts: 325
Credit: 28,109,066
RAC: 82
United States
Message 868262 - Posted: 22 Feb 2009, 22:25:56 UTC - in response to Message 868187.  

I note it in nearly all 'panic threads'.. ;-)

Rise the WUs in your cache and you will 'nearly' never run out of work..


I've got those numbers set about as high as they can go, and my queue ran dry this morning. Since adding CUDA I'm burning through WUs at a prodigeous rate.

ID: 868262 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 868263 - Posted: 22 Feb 2009, 22:32:22 UTC - in response to Message 868262.  

I've got those numbers set about as high as they can go, and my queue ran dry this morning.

I have a 4 day cache & have only run out of work about twice in the last 5 years.

Grant
Darwin NT
ID: 868263 · Report as offensive
Profile Jack Shaftoe
Avatar

Send message
Joined: 19 Aug 04
Posts: 44
Credit: 2,343,242
RAC: 0
United States
Message 868267 - Posted: 22 Feb 2009, 22:37:56 UTC - in response to Message 868263.  
Last modified: 22 Feb 2009, 22:49:27 UTC

Someone call Buckaroo Banzai and the Hong Kong Cavaliers! Maybe they can bring SETI their Oscillation Overthruster!

ID: 868267 · Report as offensive
Zydor

Send message
Joined: 4 Oct 03
Posts: 172
Credit: 491,111
RAC: 0
United Kingdom
Message 868270 - Posted: 22 Feb 2009, 22:43:00 UTC - in response to Message 868263.  

I have a 4 day cache & have only run out of work about twice in the last 5 years.


CUDA GPUs eat WUs like there is no tomorrow ..... the Well soon starts to run dry when there are infrastructure issues, due to the limit on cached CUDA WUs.
ID: 868270 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 868273 - Posted: 22 Feb 2009, 22:44:41 UTC - in response to Message 868270.  

I have a 4 day cache & have only run out of work about twice in the last 5 years.


CUDA GPUs eat WUs like there is no tomorrow ..... the Well soon starts to run dry when there are infrastructure issues, due to the limit on cached CUDA WUs.

4 days is 4 days.
Grant
Darwin NT
ID: 868273 · Report as offensive
Zydor

Send message
Joined: 4 Oct 03
Posts: 172
Credit: 491,111
RAC: 0
United Kingdom
Message 868275 - Posted: 22 Feb 2009, 22:47:12 UTC - in response to Message 868273.  

Not with CUDA its not - the CUDA cache is finite dependent on the hardware, it bares no relation to "X Days" set in the Manager app
ID: 868275 · Report as offensive
john deneer
Volunteer tester
Avatar

Send message
Joined: 16 Nov 06
Posts: 331
Credit: 20,996,606
RAC: 0
Netherlands
Message 868276 - Posted: 22 Feb 2009, 22:48:54 UTC - in response to Message 868248.  

I realize that bandwidth is maxed, that was not my point.


Looking at the server status page and Scarecrow graphs makes me wonder.

Normally MB WUs return at the rate of 10 to 20 to 1 AP WU.

Why does it suddenly appear easier to return an AP work unit than an MB WU ??

Presently the ratio is 2 to 1, prox.


The most likely reason for this is given in some other message thread (server panic thread). It seems that people have gotten into trouble downloading the astropulse executable, it is not provided by the seti servers as it should.

People download some astropulse 5.03 units, which then promptly fail (because the executable needed to crunch them could no be downloaded). The units are returned and the system asks for new units .... A very fast way of 'processing' astropulse :-) Only problem is that all of those units erroring out result in more downloads, and more, and more .....

Regards,
John.
ID: 868276 · Report as offensive
Fred W
Volunteer tester

Send message
Joined: 13 Jun 99
Posts: 2524
Credit: 11,954,210
RAC: 0
United Kingdom
Message 868277 - Posted: 22 Feb 2009, 22:49:44 UTC - in response to Message 868273.  

I have a 4 day cache & have only run out of work about twice in the last 5 years.


CUDA GPUs eat WUs like there is no tomorrow ..... the Well soon starts to run dry when there are infrastructure issues, due to the limit on cached CUDA WUs.

4 days is 4 days.

Not when Boinc estimates that every 10 minute CUDA WU will take 12.5 hours (I was testing the Boinc 6.x.x - and am now grateful for the max daily download that it dumped on me a couple of days ago :)

F.
ID: 868277 · Report as offensive
Zydor

Send message
Joined: 4 Oct 03
Posts: 172
Credit: 491,111
RAC: 0
United Kingdom
Message 868280 - Posted: 22 Feb 2009, 22:59:54 UTC - in response to Message 868277.  

I was testing the Boinc 6.x.x - and am now grateful for the max daily download that it dumped on me a couple of days ago :)


My 9800GTX meanders along at a rate of about 50-60 WUs per half day, so I am not feeling the pain as much as others as the GPU is only 50% to SETI. (I split the time half and half with GPUGRID). I dread to think how many WUs your twin 295s are eating rolf :)
ID: 868280 · Report as offensive
Marc Braun

Send message
Joined: 1 Jun 99
Posts: 2
Credit: 8,560,086
RAC: 0
Luxembourg
Message 868286 - Posted: 22 Feb 2009, 23:24:58 UTC

Now this takes some time that no uploads are accepted. Is there some issue to this, or is it a known problem?
ID: 868286 · Report as offensive
Fred W
Volunteer tester

Send message
Joined: 13 Jun 99
Posts: 2524
Credit: 11,954,210
RAC: 0
United Kingdom
Message 868289 - Posted: 22 Feb 2009, 23:37:07 UTC - in response to Message 868280.  

I was testing the Boinc 6.x.x - and am now grateful for the max daily download that it dumped on me a couple of days ago :)


My 9800GTX meanders along at a rate of about 50-60 WUs per half day, so I am not feeling the pain as much as others as the GPU is only 50% to SETI. (I split the time half and half with GPUGRID). I dread to think how many WUs your twin 295s are eating rolf :)

I've currently got about 200 CUDA's trying to upload (they can't because network access is switched off) and 12 AP's as well. Fortunately most of the AP's still to be done are 5.03's so they will take a little longer than the 5.00's. BTW, it's only one 295 (using both cores running at Stock speed - ATM :)

F.
ID: 868289 · Report as offensive
Profile Graeme of Boinc UK

Send message
Joined: 25 Nov 02
Posts: 114
Credit: 1,250,273
RAC: 0
United Kingdom
Message 868290 - Posted: 22 Feb 2009, 23:37:23 UTC
Last modified: 22 Feb 2009, 23:37:47 UTC

I suppose that someone will have to lift the panic button back up!

Normal service will likely be restored not long after 15:00 GMT Monday!
If you are really concerned about losing your standing in the race for most work done then may I suggest that you sign up for one of the other Boinc projects that exist out there on a temporary basis, i.e. set your local parameters to 1.0 days of work and be useful to some of the other Boinc projects that would really appreciate your albeit temporary crunching.

Do your homework first on which projects have short turnaround times then follow these instructions........

In Boinc manager click "tools" then click "attach to project" then pick your chosen Boinc addiction.

When Seti starts talking again then either dump the workunits of that project or run with them until they are exhausted and carry on as before on Seti.

Graeme.
ID: 868290 · Report as offensive
Fred W
Volunteer tester

Send message
Joined: 13 Jun 99
Posts: 2524
Credit: 11,954,210
RAC: 0
United Kingdom
Message 868295 - Posted: 22 Feb 2009, 23:41:17 UTC - in response to Message 868290.  

I suppose that someone will have to lift the panic button back up!


Aww - c'mon... A good panic now and again never did anyone any harm. Clears the system - a bit like an enema really (Now there's a thought for the "pipes" at Berkeley :)

F.
ID: 868295 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 14 · Next

Message boards : Number crunching : Panic Mode On (13) 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.