I have one machine that gets a lot of these errors as well. It has 3 GPU's and If I try to run more that ~18 WU's simultaneously it starts producing these errors. At 21 WU's, I see 5-15 errors per day and the number goes up exponentially from there. If I try to run 36, almost all of them error out. I haven't been able to trace the source of these errors. I know others who run 24+ WU's at a time with success so I'm guessing it's hardware related or possibly a problem with some non-BOINC software.

The latest results status information for the last error received is

Workunit Status

Project Name: Help Conquer Cancer
Created: 04/07/2013 16:11:26
Name: X0930129771353201108241026
Minimum Quorum: 2
Replication: 2


Result Name App Version Number Status Sent Time Time Due /
Return Time CPU Time / Elapsed Time (hours) Claimed/ Granted BOINC Credit
X0930129771353201108241026_ 2-- - In Progress 4/9/13 16:22:08 4/16/13 16:22:08 0.00 0.0 / 0.0
X0930129771353201108241026_ 0-- 705 Pending Validation 4/8/13 04:49:13 4/9/13 03:06:12 0.12 63.3 / 0.0
X0930129771353201108241026_ 1-- 705 Error 4/8/13 04:49:10 4/9/13 16:22:00 0.00 0.1 / 0.0

dkt,

Error code 233 occurs when the application does a performance check at the beginning of execution to make sure the video card can run the application. However with this error you should get information in stderr about the card and the time for the performance check, not sure why this is not showing up in your stderr. The error can occur if your gpu is loaded with other activity or if you try and run to many workunits at a time.

Thanks,
armstrdj

My error messages have changed, but I will have to wait a day or two to see if I am still getting the previous errors. Now, toward the end of the stream for the Error, I am getting


Result Log

Result Name: X0930130721179201111171053_ 0--
<core_client_version>7.0.28</core_client_version>
<![CDATA[
<message>
No process is on the other end of the pipe. (0xe9) - exit code 233 (0xe9)
</message>
<stderr_txt>
Commandline: projects/www.worldcommunitygrid.org
... [and a lot of lines followed by...]
Estimated kernel execution time = 3.21459 [sec]
ERROR: Kernel execution time estimate too high, exiting.
19:45:12 (10156): called boinc_finish

</stderr_txt>
]]>

I will give an update on Monday with the weekend processing results.

The last error was (sorted by return time)

X0930130720623201111171101_ 1-- SSTsComp Error 4/10/13 14:46:02 4/11/13 16:06:17 0.00 / 0.00 0.1 / 0.0

During the beta testing I remember reading that running more than 1 workunit per CPU would cause errors. Lowering the WU number to the actual # of CPU cores should clear the errors.