Condor: View Results

How to view the results of your condor job

When your jobs running in Condor finish, you are sent email that tells you about your completed job. A typical email message from condor looks like:

Your condor job
/cae/apps/bin/matlab
exited with status 0.

Submitted at: Tue Jun 8 15:58:20 2007
Completed at: Tue Jun 8 15:58:54 2007
Real Time: 0 00:00:34

Remote User Time: 0 00:00:01
Remote System Time: 0 00:00:01
Total Remote Time: 0 00:00:02

Local User Time: 0 00:00:00
Local System Time: 0 00:00:00
Total Local Time: 0 00:00:00

Virtual Image Size: 1 Kilobytes

The meanings of the above lines are:
Exited with status n The exit value of the program. Most programs exit with 0 if everything went well. If things did not go well, a different number is used as the exit value. If the value is greater than 128, the program was killed as the result of receiving a signal. In general a value greater than 128 is an indication that there are problems within the program. See "man 7 signal" or /usr/include/sys/signal.h for a list of signal numbers and a short description. For example, if you had some problems with pointer arithmetic, you may get a bus error (SIGBUS). The exit value would be 138.
Submitted at: The time that you ran condor_submit.
Completed at: The time that the job completed. This should be the same time that the mail message was sent to you.
Real time: The amount of time that elapsed between submitted at and completed at. It has the format of D HH:MM:SS (days hours:minutes:seconds).
Remote user time: The amount of time that your executable spent on the CPU in user mode. User mode operations include pretty much everything other than I/O. In general, user mode should make up the bulk of your CPU time.
Remote system time: The amount of time that your executable spent on the CPU in system mode. System mode time consists of time that it spent in system calls such as open(), close(), read(), write(), stat(), etc.
Total remote time: The sum of remote user time and remote system time.
Local user time: Amount of user time consumed by the condor_shadow process on the machine from which the job was submitted. Note that if you run condor_submit from sun-2, sun-2 runs a "helper" script that causes the job to be really submitted from spacewar.cae.wisc.edu.
Local system time: Amount of system time consumed by the condor_shadow process on the machine from which the job was submitted. Note that if you run condor_submit from sun-2, sun-2 runs a "helper" script that causes the job to be really submitted from spacewar.cae.wisc.edu.
Total local time: Sum of local user time and local system time.
Virtual image size: Amount of memory used by the job.

Depending on how your submit description file was written and how your job normally would run, you should check the files defined by the variables Output and Error.




Keywords:email log   Doc ID:6742
Owner:Ken H.Group:Computer-Aided Engineering
Created:2007-11-20 19:00 CDTUpdated:2016-07-01 15:15 CDT
Sites:Computer-Aided Engineering
Feedback:  0   0