Troubleshooting a failing service

[sidebrnz]

For several months, now, I've been using dynamic DNS from https://www.dnsexit.com/ to make my home box visible on the Internet as a subdomain of my own private domain. Everything worked fine until I "upgraded" to F16 and the New! Improved! Shiny! systemd. Now, I get an error message at boot that the service failed and if I run

systemctl status ipUpdate.service

I'm told that it failed with this error: (code=exited, status=203/EXEC). And, if I use nslookup, I get the wrong IP. However, if I run this:

/usr/sbin/ipUpdate.pl

it works Just Fine, and I start getting the right IP. (From what I can tell, running that script is the core of what the service does.) I've emailed their support twice with a request for assistance but received no reply because they only support paying customers and my domain's not hosted there. Does anybody know how to troubleshoot this? If not, I'll probably have to set it up as a cron job, running the script directly but I'd rather not go that route if I don't have to.

[jpollard]

It may be running the "service" before the network is ready. This is a normal problem of systemd as it does not easily permit scheduling of startup processes.

You can try adding an "after=NetworkManager-wait-online.service" and possibly adding "network.target" as another.

[sidebrnz]

The service runs every few hours so that if my dynamic IP changes, my DNS gets updated. No matter when I check, it tells me the same thing:

[root@khorlia Desktop]# systemctl status ipUpdate.service
ipUpdate.service - LSB: DnsExit.Com dynamic dns client.
Loaded: loaded (/etc/rc.d/init.d/ipUpdate)
Active: failed since Tue, 14 Feb 2012 18:28:00 -0800; 19h ago
CGroup: name=systemd:/system/ipUpdate.service

As I only reboot for a kernel update, and haven't had time to do so since the one this morning, you can see what I mean. And, here's what I get if I stop and restart it:

[root@khorlia Desktop]# systemctl stop ipUpdate.service
[root@khorlia Desktop]# systemctl start ipUpdate.service
Job failed. See system logs and 'systemctl status' for details.
[root@khorlia Desktop]# systemctl status ipUpdate.service
ipUpdate.service - LSB: DnsExit.Com dynamic dns client.
Loaded: loaded (/etc/rc.d/init.d/ipUpdate)
Active: failed since Wed, 15 Feb 2012 14:11:59 -0800; 12s ago
Process: 21010 ExecStart=/etc/rc.d/init.d/ipUpdate start (code=exited, status=203/EXEC)
CGroup: name=systemd:/system/ipUpdate.service

[jpollard]

And what is in the /system/ipUpdate.service file?

[sidebrnz]

Here's the contents of /etc/rc.d/init.d/ipUpdate, which worked Just Fine under F14:

# chkconfig: 345 85 60
# description: DnsExit.Com dynamic dns client.
# processname: ipUpdate.pl

### BEGIN INIT INFO
# Provides: ipUpdate
# Required-Start: $local_fs $network 
# Should-Start: 
# Required-Stop:
# Default-Start: 3 4 5
# Default-Stop: 0 1 2 6
# Short-Description: DnsExit.Com dynamic dns client.
# Description: DnsExit.Com dynamic dns client.
### END INIT INFO

# Source function library.
if [ -f /etc/SuSE-release ]; then
   SUSE=1
   START=startproc
   STATUS="rc_status -v"
   . /etc/rc.status
   rc_reset
else
   SUSE=0
   START=daemon
   STATUS=echo
   . /etc/rc.d/init.d/functions

fi

start() {
	echo -n $"Starting ipUpdate: "
	$START /usr/sbin/ipUpdate.pl
	touch /var/lock/subsys/ipUpdate
	$STATUS
}

stop() {
	echo -n $"Shutting down ipUpdate: "
	killproc ipUpdate.pl
	rm -f  /var/lock/subsys/ipUpdate
        $STATUS
}

[ -f /usr/sbin/ipUpdate.pl ] || exit 0
grep "daemon[:space:]*=[:space:]*no" /etc/dnsexit.conf >/dev/null 2>&1 && exit 0
case "$1" in
    start)
        start
        ;;
    stop)
        stop
        ;;
    status)
        ;;
    restart)
        stop
        start
        ;;
    condrestart)
        ;;
    reload)
        ;;
    *)
        echo $"Usage: $0 {start|stop|status|restart|condrestart|reload}"
        ;;
esac
[ $SUSE -eq "1" ] && rc_exit
exit 0

Please note that there doesn't seem to be anything in the file that generates that kind of error message, and I don't know enough about this kind of thing to work our what's causing it now that wasn't doing it before.

[cazo]

It looks to me like the code is missing something

else
   SUSE=0
   START=daemon
   STATUS=echo
   . /etc/rc.d/init.d/functions
fi

start() {
	echo -n $"Starting ipUpdate: "
	$START /usr/sbin/ipUpdate.pl
	touch /var/lock/subsys/ipUpdate
	$STATUS
}

I think that just calls echo (after running 'ipUpdate.pl'), instead of returning anything useful. You can try adding something like:

else
   SUSE=0
   START=daemon
   STATUS=echo
   . /etc/rc.d/init.d/functions
fi

start() {
	echo -n $"Starting ipUpdate: "
	$START /usr/sbin/ipUpdate.pl
        RETVAL=$?
	touch /var/lock/subsys/ipUpdate
	$STATUS
        return $RETVAL
}

and see if that helps.

[sidebrnz]

Remember, this script worked Just Fine, TYVM, exactly as it is now, right up until I "upgraded" to F16.

[jpollard]

F16 uses systemd which is NOT a SysVinit compatible system, so things don't work the same.

I expect you put a file in /lib/systemd/system (or /etc/systemd/system) named ipUpdate.service.

This file determines how/when it runs. What does this file contain?

Or are you trying to use the SysVinit "compatability" operation?

[sidebrnz]

I installed the service (from a .rpm) when I was running F14 and it worked. Now that I've "upgraded" to F16, it doesn't. There's no file in either of those two directories that appears to control this, so I have no idea what's calling it or how.

[jpollard]

Try moving the startup into rc.local, and change the options on rc.local to delay until after the network is functional.

There are a number of threads on that subject.

[sidebrnz]

I don't know if I haven't made myself clear, or if you're just not reading what I've written.

If I were to restart that service RIGHT NOW, without rebooting, it would still fail in exactly the same way. I know, because I've tried it, more than once. Changing the start time at boot won't help.