I have a job I need to run automatically at about every 5 minutes. Cron could certainly do that. But if the job should run long, I don't want it to be started again while a previous instance is still running. And if the job should run 4.5 minutes, I don't necessarily want it to run again in 30 seconds, tho that wouldn't be fatal.
I could just use a script with a delay at the end before it loops back to the top, but that's crude and not particularly reliable.
Any clever, robust way to do this?
with cron & an a more intelligent script :)
seriously, you can have your script create some kind of lock file, and then whenever it runs, check for the existence of the lock file before proceeding. if it doesn't find one, then it should create a new lock file, do its thing, and when completed, delete its lock file.
if you don't want it to do anything if its been more than a specific since the last one finished, you'll need to add some time/date stamp analysis functionality to the script so that it can determine how long its been since the last job ran. perhaps creating both a lock file, which is transient, and another file that just gets the output from 'date' cat'ed into it when the last job finishes.
-- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ L. Friedman [EMAIL PROTECTED] Linux Step-by-step & TyGeMo: http://netllama.ipfox.com
1:25pm up 1 day, 19:53, 2 users, load average: 0.01, 0.11, 0.12
_______________________________________________ Linux-users mailing list [EMAIL PROTECTED] Unsubscribe/Suspend/Etc -> http://www.linux-sxs.org/mailman/listinfo/linux-users
