We use a queueing system to submit thousands of jobs in a Linux environment (LSF job arrays). It is necessary to periodically check the status of the jobs. However, I have been unable to coax the status command to provide meaningful names for the jobs, rather than just their job ID number (for example, if I submit 5000 jobs, I want to know if job "foo" and job "bar" are still running, instead of ID647 and ID4902).
To do this, I would have to sift through the vebose and poorly-formatted output of the status command to find the directories holding the output status files. Then I'd have to grep throught the output files to correlate ID's with job names. Bor-ing.
Within a few hours, I cranked out some decent Perl code to do the job. But, I must admit that hanging out at PerlMonks for the last couple years has given me the confidence and skills to rapidly converge on and implement a solution. A year and a half ago, I probably would have thrashed around for a few days or a week on that task, or perhaps I would have found the task too daunting.
Had I never joined, this...
Instead, what I have...
I guess this is as much a PerlMonks testimonial as it is a Perl testimonial. Without Perl, I'd be in a world of hurt. Without the Monks, I'd have a really embarrassing heap of unmaintainable, unreadable, undocumented, buggy code!
In reply to Re: How has Perl(Monks) saved you time at your job this week|month|year|decade?
by toolic
in thread How has Perl saved you time at your job this week|month|year|decade?
by missingthepoint
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |