> Hello, all:
>
> I am trying to create a job which will run for a long time, so I need to
> make sure that it:
>
> * starts up automatically
>
> * is always running
>
> * checks to see if it is running and restarts itself if the previous
> instance died
>
> * never runs two copies of itself
>
> I have read through the files in /etc/rc.d/init.d for hints, since I know
> that somehow they put the PID in a file and then check that file to see if
> it's running or not. Duplicating this functionality is one possibility.
>
> I am running a shell script that runs an rsync process, so perhaps even
> something as simple as checking the exit status of rsync with a while loop
> inside the script would work. However, I don't know how to do that, and my
> reading of the bash man page has not provided enough enlightenment.
>
> Anyone care to offer suggestions?
Create a script called run_application_ like this:
#!/bin/sh
while :
do
_application_
sleep 10
done
(Replace _application_ with the name of the application).
Then in /etc/rc.d/init.d create a startup script that calls the
run_application_ script and make links in the required runlevels.
That will start the job and if it dies wait 10 seconds and start it
again. You could add some features:
logging starts and stops
keep a count of stops and starts and if it fails too often break
out of the loop and mail a warning
keep a count of stops and starts and if it fails too frequently in
a short space of time break out of the loop and mail a warning
--
Ian
--
redhat-list mailing list
unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe
https://www.redhat.com/mailman/listinfo/redhat-list
