Check if script running ... if not, start it.

I'm trying to do this in sh:

I want to set up a cron job that periodically will run a check.sh script to see if a PHP script called proxy.php is running on the webserver. If it is, do nothing ... if it's not, start it up. I'm lost as to how to go about determing if the script is running. Any help you can provide is really appreciated by this newbie!

This post has been edited 1 time(s), it was last edited by forefront on 06-23-2003 at 22:09.

Hmm, I can't say that I've ever done something like that before. I'll look into it.

Why does it need to be a cronjob?

Utinam logica falsa tuam philosophiam totam suffodiant

This post has been edited 1 time(s), it was last edited by Heather on 06-24-2003 at 00:25.

The cron will do the periodic check running the basic script. If the script has no argument (i.e. start, stop or restart), then it just checks to see if the program is running. If not, it starts it.

The proxy.php is an anti-spam system that sits between the POP3 email client and the standard POP3 service.

The software works, but is still being debugged and further tested. Sometimes it stops and we're still trying to find out why, but if it stops, users cannot get mail without reconfiguring their email clients.

I think you have to make a script to do the functionality first and then just call that script in cronjob. Sorry, but I can't help you with the script but I think that's how you should do it.

Forefront...cron jobs don't do anything but run commands at pre-assigned times. What you first need to do is make a script that does what you need it to do. It can be cgi, sh, or anything else. After that is done and it works, then you set it up to be run in a cron job.

Utinam logica falsa tuam philosophiam totam suffodiant

Uh ... I guess I wasn't very clear. I know I need to write a script and I'm using sh or maybe bash. I know how to search to see if a script or program is running, but I don't know what happens when it's not running, how to parse the result, detect the response and have my script respond appropriately. I've searched high and low on the Net for anything similar, but so far I've come up blank.

This post has been edited 1 time(s), it was last edited by forefront on 06-24-2003 at 20:19.

quote:

Originally posted by forefront Uh ... I guess I wasn't very clear. I know I need to write a script and I'm using sh or maybe bash. I know how to search to see if a script or program is running, but I don't know what happens when it's not running, how to parse the result, detect the response and have my script respond appropriately. I've searched high and low on the Net for anything similar, but so far I've come up blank.

Well, parsing a script's progress...I know how to make it print progress indictators into an ssh session...but I'm not 100% sure what you're asking.

Utinam logica falsa tuam philosophiam totam suffodiant

When the script is run by NOT run by crontab, it will contain normally contain an argument of stop, start, or restart. When crontab runs it, those arguments will be missing, so the script logic will be to check to see if proxy.php is currently running and if it is, do nothing and quit ... otherwise do the 'start' function.

For example, here's the 'start' function:
start()
{
command="nohup proxy.php &"
exit 0
}

... and the 'stop' function:
stop()
{
command="ps -auxww | grep \"php\" | awk '{print $2}' | xargs kill -9"
exit 0
}

So, if I want to have a check function, I get this far:
check()
{
command="ps -auxww | grep \"php\"" | awk '{print $2}'
}
This returns the PID. If it is not running, it returns nothing. How do I detect nothing? Once I know how to do that, then I can start/exit or just exit. It gets easy then.

This post has been edited 3 time(s), it was last edited by forefront on 06-24-2003 at 23:49.

I don't know anything about the language you are using but I would imagine that if you want to detect nothing you have to detect if it is something else than the pid. I don't know if it works but that's the logic I would probably use.

I'm down to one basic problem to make this work:

How do I get the results of the "command=..." into a variable?

Once that's done, I can test to see if it's null or not.