Haunted by the 197 (0xc5) EXIT_TIME_LIMIT_EXCEEDED error

Message boards : Number crunching : Haunted by the 197 (0xc5) EXIT_TIME_LIMIT_EXCEEDED error
Message board moderation

To post messages, you must log in.

AuthorMessage
Ivailo Bonev
Volunteer tester
Avatar

Send message
Joined: 26 Jun 00
Posts: 247
Credit: 35,864,461
RAC: 2
Bulgaria
Message 1347543 - Posted: 17 Mar 2013, 5:01:07 UTC
Last modified: 17 Mar 2013, 5:19:25 UTC

Something strange has happened to my PC today. After strange errors from the NVidia driver, I decided to install newest one (and updated BOINC to 7.0.56), but all my CUDA tasks continue to error out with EXIT_TIME_LIMIT_EXCEEDED error, and I see crazy times on my GF9800GT, like 1min (normally tasks are 15-25min).
http://setiathome.berkeley.edu/result.php?resultid=2876991557
It's for the first time, and I don't know what to do. Do I need to abort all cuda tasks?
ID: 1347543 · Report as offensive
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 22160
Credit: 416,307,556
RAC: 380
United Kingdom
Message 1347562 - Posted: 17 Mar 2013, 6:48:51 UTC

Did you just update BOINC to one of the newer versions?

I had the same happen when I moved from 7.0.28 to 7.0.56 the other day. My solution was to roll back to 7.0.28 and expected durations went back to the "normal" values.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1347562 · Report as offensive
Profile Bernie Vine
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 26 May 99
Posts: 9954
Credit: 103,452,613
RAC: 328
United Kingdom
Message 1347580 - Posted: 17 Mar 2013, 9:06:22 UTC

In will repost Claggys answer to me when I had the same problem:
This is because of a change in Boinc 7.0.33, the flops value for GPU tasks gets increased by by times 10 (when no flops values have been supplied in the app_info), existing GPU tasks will be put on the verge of going Maximum time exceeded:

client: when estimating FLOPS for an anonymous-platform app version for which no estimate has been supplied by user, use (CPU speed)*(cpu_usage + 10*gpu_usage) (--> add the 10*)


Fresh GPU tasks will get revised <rsc_fpops_est> and <rsc_fpops_bound> values and will complete O.K, you can eithier reset the project and get your tasks resent,
or use the -177 protection option in BoincRecheduler (but that'll mess up the time estimates)


I reset the project and everything was OK after that.
ID: 1347580 · Report as offensive
Ivailo Bonev
Volunteer tester
Avatar

Send message
Joined: 26 Jun 00
Posts: 247
Credit: 35,864,461
RAC: 2
Bulgaria
Message 1347687 - Posted: 17 Mar 2013, 13:43:59 UTC

Thanks, reseting the project. I hope new tasks will be ok.
ID: 1347687 · Report as offensive
Profile Bernie Vine
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 26 May 99
Posts: 9954
Credit: 103,452,613
RAC: 328
United Kingdom
Message 1347722 - Posted: 17 Mar 2013, 15:00:38 UTC - in response to Message 1347687.  

Thanks, reseting the project. I hope new tasks will be ok.

They should be, it is because you were trying to complete GPU tasks with the new version of BOINC that had been started in a version prior to 7.0.33

Anyone else reading this who hasn't upgraded above BOINC 7.0.33, and who has no flops values supplied in the app_info, set NNT and run down you GPU tasks before upgrading, after which all new GPU task will be OK
ID: 1347722 · Report as offensive
Ivailo Bonev
Volunteer tester
Avatar

Send message
Joined: 26 Jun 00
Posts: 247
Credit: 35,864,461
RAC: 2
Bulgaria
Message 1348012 - Posted: 18 Mar 2013, 7:42:52 UTC

It's working normally now, thanks again. :)
ID: 1348012 · Report as offensive
Highlander
Avatar

Send message
Joined: 5 Oct 99
Posts: 167
Credit: 37,987,668
RAC: 16
Germany
Message 1348039 - Posted: 18 Mar 2013, 11:26:55 UTC

Vote for sticky-thread
- Performance is not a simple linear function of the number of CPUs you throw at the problem. -
ID: 1348039 · Report as offensive
Profile William
Volunteer tester
Avatar

Send message
Joined: 14 Feb 13
Posts: 2037
Credit: 17,689,662
RAC: 0
Message 1348060 - Posted: 18 Mar 2013, 13:18:32 UTC - in response to Message 1348039.  

Vote for sticky-thread

So schnell schiessen die Preussen nicht!

IOW we're working on it...
A person who won't read has no advantage over one who can't read. (Mark Twain)
ID: 1348060 · Report as offensive

Message boards : Number crunching : Haunted by the 197 (0xc5) EXIT_TIME_LIMIT_EXCEEDED error


 
©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.