OK...how did THIS timeout happen?


log in

Advanced search

Message boards : Number crunching : OK...how did THIS timeout happen?

Author Message
msattler
Volunteer tester
Avatar
Send message
Joined: 9 Jul 00
Posts: 38154
Credit: 556,300,858
RAC: 593,145
United States
Message 1321095 - Posted: 29 Dec 2012, 2:38:24 UTC

AP WU 1142265986.
Sent to one of my rig and one other on 28 Dec 2012 | 22:23:40 UTC.

The other user has a deadline of 22 Jan 2013 | 22:23:40 UTC.
Mine shows time out, no response at 29 Dec 2012 | 2:28:26 UTC.
It was then reissued with a deadline of 23 Jan 2013 | 2:28:37 UTC.

Why didn't the kitties have until Jan. 22nd to finish it?
____________
*********************************************
Embrace your inner kitty...ya know ya wanna!

I have met a few friends in my life.
Most were cats.

Profile Wiggo
Avatar
Send message
Joined: 24 Jan 00
Posts: 6452
Credit: 90,126,634
RAC: 72,374
Australia
Message 1321099 - Posted: 29 Dec 2012, 2:44:29 UTC - in response to Message 1321095.

I've always thought that it was part of the same mechanism that stops VLAR's being sent on a nVIDIA request myself.

Cheers.
____________

jravin
Send message
Joined: 25 Mar 02
Posts: 927
Credit: 94,525,999
RAC: 83,660
United States
Message 1321106 - Posted: 29 Dec 2012, 3:38:52 UTC - in response to Message 1321099.

I've always thought that it was part of the same mechanism that stops VLAR's being sent on a nVIDIA request myself.


Could you please explain this remark? Is it a joke? Am I missing something (besides 100 or so IQ points)?

____________

msattler
Volunteer tester
Avatar
Send message
Joined: 9 Jul 00
Posts: 38154
Credit: 556,300,858
RAC: 593,145
United States
Message 1321120 - Posted: 29 Dec 2012, 4:30:15 UTC

Got another one just about like it.......must be the idiot scheduler sending WUs where they don't belong because comms are trashed.
____________
*********************************************
Embrace your inner kitty...ya know ya wanna!

I have met a few friends in my life.
Most were cats.

Profile Wiggo
Avatar
Send message
Joined: 24 Jan 00
Posts: 6452
Credit: 90,126,634
RAC: 72,374
Australia
Message 1321145 - Posted: 29 Dec 2012, 5:49:55 UTC - in response to Message 1321106.

I've always thought that it was part of the same mechanism that stops VLAR's being sent on a nVIDIA request myself.


Could you please explain this remark? Is it a joke? Am I missing something (besides 100 or so IQ points)?

With people who are not running the nVIDIA openCL app then the Astropulse workunit will be treated the same as a VLAR, CPU request makes it through but isn't acknowledged at the user end so the workunit becomes a ghost and then the machine requests GPU work so any ghosts that can only run on CPU for that setup will be automatically timed out.

Cheers.
____________

Message boards : Number crunching : OK...how did THIS timeout happen?

Copyright © 2014 University of California