Panic Mode On (80) Server Problems?


log in

Advanced search

Message boards : Number crunching : Panic Mode On (80) Server Problems?

Previous · 1 . . . 22 · 23 · 24 · 25
Author Message
Grant (SSSF)
Send message
Joined: 19 Aug 99
Posts: 5685
Credit: 56,113,625
RAC: 49,727
Australia
Message 1332693 - Posted: 30 Jan 2013, 7:35:35 UTC - in response to Message 1332690.

Also, it looks like MB is about to run out of tapes to split.

Will probably last for a while as there is only on MB splitter that is alive at the moment. But it probably won't be long before it is needed- that huge Ready to Send buffer is shrinking rapidly.

____________
Grant
Darwin NT.

BetelgeuseFive
Volunteer tester
Send message
Joined: 6 Jul 99
Posts: 56
Credit: 4,843,568
RAC: 9,453
Netherlands
Message 1332819 - Posted: 30 Jan 2013, 15:41:08 UTC


Hello folks,

What is the limit for a VLAR task ?
I received a number of tasks that take much longer than usual on my GPU.
Angle for these tasks seems to be 0.125883

http://setiathome.berkeley.edu/result.php?resultid=2816144089

Are VLAR tasks being sent to GPU or are these tasks just closer to the limit than anything I have seen before ?

Tom

P.S. Sorry I posted this in the wrong thread first

____________

rob smith
Volunteer tester
Send message
Joined: 7 Mar 03
Posts: 8122
Credit: 52,313,347
RAC: 78,043
United Kingdom
Message 1332834 - Posted: 30 Jan 2013, 16:48:47 UTC

Do you mean an angular limit or an execution time limit?
____________
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

BetelgeuseFive
Volunteer tester
Send message
Joined: 6 Jul 99
Posts: 56
Credit: 4,843,568
RAC: 9,453
Netherlands
Message 1332839 - Posted: 30 Jan 2013, 16:58:04 UTC - in response to Message 1332834.

Do you mean an angular limit or an execution time limit?


I mean the angle. I didn't know there was a difference in handling execution time between VLAR and non-VLAR tasks. I thought the maximum execution time was always the estimated time multiplied by a constant (although I forgot the value of the constant).

Tom

____________

Josef W. Segur
Volunteer developer
Volunteer tester
Send message
Joined: 30 Oct 99
Posts: 4196
Credit: 1,028,990
RAC: 265
United States
Message 1332876 - Posted: 30 Jan 2013, 18:07:30 UTC - in response to Message 1332839.

Do you mean an angular limit or an execution time limit?


I mean the angle. I didn't know there was a difference in handling execution time between VLAR and non-VLAR tasks. I thought the maximum execution time was always the estimated time multiplied by a constant (although I forgot the value of the constant).

Tom

MB tasks with AR < 0.120000 get the .vlar extension.

Maximum execution time is indeed directly related to the raw estimate, a 10x multiple. However, the raw estimate is multiplied by Duration Correction Factor to give the actual estimate but the time limit does not use DCF.
Joe

Profile KWSN Ekky Ekky Ekky
Avatar
Send message
Joined: 25 May 99
Posts: 922
Credit: 10,844,951
RAC: 12,785
United Kingdom
Message 1333489 - Posted: 1 Feb 2013, 10:02:29 UTC

The thin blue line on Cricket is heading south. More reporting problems. I'm sure it will all be fixed when they get in for work but this could be ominous coming before the weekend.
____________

Claggy
Volunteer tester
Send message
Joined: 5 Jul 99
Posts: 4039
Credit: 32,691,382
RAC: 712
United Kingdom
Message 1333490 - Posted: 1 Feb 2013, 10:07:35 UTC
Last modified: 1 Feb 2013, 10:12:14 UTC

Data Out on the Cricket Graph has dropped almost to the floor, and scheduler contacts now can't connect to reference site.

Graphs for gigabitethernet2_3

Claggy

Profile Link
Avatar
Send message
Joined: 18 Sep 03
Posts: 823
Credit: 1,544,595
RAC: 470
Germany
Message 1333491 - Posted: 1 Feb 2013, 10:13:18 UTC - in response to Message 1333490.

scheduler contacts now can't connect to reference site.

Can't connect to reference site means you can't connect to google.com (or whatever you have specified in your cc_config) which has very likely nothing to do with what you see on the graph.
____________
.

Claggy
Volunteer tester
Send message
Joined: 5 Jul 99
Posts: 4039
Credit: 32,691,382
RAC: 712
United Kingdom
Message 1333501 - Posted: 1 Feb 2013, 11:30:33 UTC - in response to Message 1333491.

scheduler contacts now can't connect to reference site.

Can't connect to reference site means you can't connect to google.com (or whatever you have specified in your cc_config) which has very likely nothing to do with what you see on the graph.


Actually, it was a attempting access to reference site, and not that the reference site was down, i misread:

01/02/2013 10:03:38 Project communication failed: attempting access to reference site
01/02/2013 10:03:38 SETI@home Scheduler request failed: Couldn't connect to server
01/02/2013 10:03:38 SETI@home [sched_op_debug] Deferring communication for 1 min 0 sec
01/02/2013 10:03:38 SETI@home [sched_op_debug] Reason: Scheduler request failed
01/02/2013 10:03:41 Internet access OK - project servers may be temporarily down.


Claggy

juan BFB
Volunteer tester
Avatar
Send message
Joined: 16 Mar 07
Posts: 4923
Credit: 267,405,543
RAC: 336,535
Brazil
Message 1333506 - Posted: 1 Feb 2013, 12:17:00 UTC
Last modified: 1 Feb 2013, 12:22:13 UTC

Sure something must be wrong in the lab, and is to early in CA to anyone be there now, maybe just the ghosts on the machines.

Requests UL/DL all end´s on a black hole.

All my hosts are running out of GPU work in the SETI dryland, soon they will be out of CPU work too.

But something calls my attention, why the cricket graph green line still shows 100% of bandwidth usage if anyone can´t UL/Dl nothing? Maybe someone or something else is taking a "free ride" on the Internet connection of the lab? That could be a source of our connection problems?
____________

Alfred and Pauline
Send message
Joined: 13 Oct 99
Posts: 25
Credit: 11,136,774
RAC: 13,701
Canada
Message 1333507 - Posted: 1 Feb 2013, 12:20:49 UTC

Stuck Astropulse V6

have ap-28mr10aa-b2-p1-00211-2012059-14928 sitting on page since 26 Jun 2012.
2 computers are validated and one pending.
Is there something I need to do to remove this or is it lost?

Profile Tim
Volunteer tester
Avatar
Send message
Joined: 19 May 99
Posts: 196
Credit: 232,041,668
RAC: 111,834
Greece
Message 1333508 - Posted: 1 Feb 2013, 12:44:22 UTC - in response to Message 1333507.

Stuck Astropulse V6

have ap-28mr10aa-b2-p1-00211-2012059-14928 sitting on page since 26 Jun 2012.
2 computers are validated and one pending.
Is there something I need to do to remove this or is it lost?


Just wait and wait and wait and…..

I have a stuck wu since Aug. 2010 :-)

Tim

____________

Profile ivan
Volunteer tester
Avatar
Send message
Joined: 5 Mar 01
Posts: 579
Credit: 130,780,527
RAC: 112,951
United Kingdom
Message 1333518 - Posted: 1 Feb 2013, 13:55:57 UTC - in response to Message 1333506.

But something calls my attention, why the cricket graph green line still shows 100% of bandwidth usage if anyone can´t UL/Dl nothing? Maybe someone or something else is taking a "free ride" on the Internet connection of the lab? That could be a source of our connection problems?

That's normal -- it takes a while for the backlog of pending downloads to be worked through, especially if it contains the large AP w/us. Looks like it's coming down now. The server status page has not been updated since about an hour after the inbound traffic started to fall, as well.
____________

Previous · 1 . . . 22 · 23 · 24 · 25

Message boards : Number crunching : Panic Mode On (80) Server Problems?

Copyright © 2014 University of California