Skip to main content

Exit codes

reference · Reviewed 2026-07-30 GA

The last thing process_output.log prints is the exit code of your launch command. These are standard shell exit codes, reported through unchanged.

Codes​

CodeMeaningUsual cause on Rescale
0No explicit errorCommand ran and exited cleanly — see the caveat below
1Catchall for general errorsSolver reported an error; read the log body
2Misuse of a shell built-inSyntax error in your launch command
126Command cannot executeA script without the execute bit — chmod +x it
127Command not foundTypo, or the software module wasn't loaded
128Invalid argument to exitRare; a script called exit with a bad value
128+nFatal error signal nThe process was killed by signal n
130Terminated by Ctrl-C (SIGINT)Job stopped from the platform
137Killed (SIGKILL)Almost always out of memory — check cluster health
255Exit status out of rangeA script returned a value above 255

Exit code 0 does not mean the run was good​

This is the single most important thing on this page. Exit code 0 means your command exited without an error status. It does not mean:

  • the solver converged,
  • the results are physically meaningful, or
  • the solver even ran — a launch script that fails silently and then exits normally still returns 0.

Always inspect the solver's own output, not just the exit code. See reading logs for which file to open.

137 is a memory problem​

If you take one diagnostic shortcut from this page: 137 means the Linux OOM killer terminated your process. The cluster ran out of memory.

Resubmit on a coretype with more memory per core, or reduce the model size. Confirm it by checking free memory in cluster health, you'll see it trending to zero before the kill.

Finding the code​

The exit code is at the very end of process_output.log, on the Results page. For a full walkthrough, see Troubleshooting Jobs by Reading process_output.log and Exit Codes.