jman fails when jobs are deleted before they are finished
Sometimes, the local scheduler is running and processing jobs. In another shell, you can delete some of the jobs that are currently running. For the local scheduler, this does not include that the jobs are actually stopped, so they still run on the machine.
Now, when killing a job (that was previously deleted) using the kill
command, jman
crashes with the following error message:
Traceback (most recent call last):
File "jman", line 26, in <module>
sys.exit(gridtk.script.jman.main())
File "gridtk/script/jman.py", line 383, in main
args.func(args)
File "gridtk/script/jman.py", line 179, in run_scheduler
jm.run_scheduler(parallel_jobs=args.parallel, job_ids=get_ids(args.job_ids), sleep_time=args.sleep_time, die_when_finished=args.die_when_finished, no_log=args.no_log_files, nice=args.nice, verbosity=args.verbose)
File "gridtk/local.py", line 191, in run_scheduler
result = "%s (%d)" % (jj.status, jj.result) if jj.result is not None else "%s (?)" % jj.status
AttributeError: 'NoneType' object has no attribute 'result'
(Note that the lines might not be up to date, I was using an older version of gridtk).