Questions and Answers :
Unix/Linux :
CPU units getting errors
Message board moderation
Author | Message |
---|---|
Send message Joined: 2 May 11 Posts: 65 Credit: 242,754,987 RAC: 0 |
when running the cpu app, the CPU units seems to fails to run properly, i'm getting: Thu 26 May 2011 07:17:52 PM EDT | Moo! Wrapper | Starting dnetc_r72_1306448940_12_12_0 Thu 26 May 2011 07:17:52 PM EDT | Moo! Wrapper | Starting task dnetc_r72_1306448940_12_12_0 using dnetc version 102 Thu 26 May 2011 07:17:54 PM EDT | Moo! Wrapper | Computation for task dnetc_r72_1306448940_12_12_0 finished Thu 26 May 2011 07:17:54 PM EDT | Moo! Wrapper | Output file dnetc_r72_1306448940_12_12_0_0 for task dnetc_r72_1306448940_12_12_0 absent This is the stderr output: http://moowrap.net/result.php?resultid=357454 |
Send message Joined: 2 May 11 Posts: 8 Credit: 503,560 RAC: 0 |
|
Send message Joined: 20 Apr 11 Posts: 388 Credit: 822,356,221 RAC: 0 |
My Linux host is failing all WUs with the same status error. Looks like another signal 11 (SIGSEGV or segmentation fault), which indicates a bug in the wrapper. I need to investigate and fix it for next app version. Until that's done, any constantly failing hosts should crunch on something else. :( (This even affects my own Linux CPU crunchers.) -w |
Send message Joined: 20 Jun 11 Posts: 2 Credit: 4,199,383 RAC: 0 |
For what it's worth (maybe nothing) I seem to be having the same issue. The first five failed at exactly 575.31 seconds of runtime, and the rest within two seconds after reaching one hour of runtime. On the other hand, the ATI workunits seem to be running quite happily so far. |
Send message Joined: 29 Jun 11 Posts: 1 Credit: 0 RAC: 0 |
Same here! Is this going to be fixed any time soon? |
Send message Joined: 24 Sep 11 Posts: 3 Credit: 34,704,414 RAC: 0 |
I get similar errors on CPU units, however only and allways when I suspend the tasks. Nvidia units suspend correctly. The tasks seem not to respond to some signals from boinc (after suspend request they still run and consume CPU, while boincmgr shows "computation error" status, I had to kill them manually). my boinc version: 6.12.33 (linux), dnetc version - 1.02 (downloaded automagicaly by boinc) |
Send message Joined: 2 May 11 Posts: 8 Credit: 503,560 RAC: 0 |
FWIW, Here's what I see in stderr.txt for this WU: [Sep 30 08:18:55 UTC] RC5-72: Completed CF:D90F55B6:00000000 (1.00 stats units) 0.00:13:35.35 - [5,267,571 keys/s] [Sep 30 08:18:55 UTC] Shutdown - packet limit exceeded. [Sep 30 08:18:55 UTC] RC5-72: Summary: 1 packet (1.00 stats units) 0.00:13:35.35 - [5,267,571 keys/s] [Sep 30 08:18:55 UTC] RC5-72: 0 packets remain in in.r72 [Sep 30 08:18:55 UTC] RC5-72: 1 packet (1.00 stats units) is in out.r72 [Sep 30 08:18:55 UTC] *Break* Shutting down... [Sep 30 08:18:55 UTC] Shutdown complete. HTH |
Send message Joined: 22 Oct 11 Posts: 4 Credit: 3,145,103 RAC: 1 |
same for me (sample result) process got signal 11 after the result has finished 7 of 8 packets. http://moowrap.net/result.php?resultid=3509487 Host (Ubuntu 11.10 64bit) http://moowrap.net/results.php?hostid=4968 on an other host (also Ubuntu 11.10 64bit) http://moowrap.net/results.php?hostid=4972 the wrapper crashed and the dnetc program is still calculating on the result. could only terminate the dnetc program due shutting down the machine. Edit Error message: 1 cruncher has been started. Matthias |
Send message Joined: 17 Jun 11 Posts: 4 Credit: 89,874,328 RAC: 33,099 |
I get similar errors on CPU units, however only and allways when I suspend the tasks. Nvidia units suspend correctly. The tasks seem not to respond to some signals from boinc (after suspend request they still run and consume CPU, while boincmgr shows "computation error" status, I had to kill them manually). I do get same error on Ubuntu CPU tasks. my boinc version: 6.12.33 (linux), dnetc version - 1.02 (downloaded automagicaly by boinc) "I always win law permanently" |