Re: Service status

Am 04.10.2015 um 18:32 schrieb Matt.:
> It's wat suits us best, si the way we are going with for about 3 years
> now (and more actually).
>
> As you can see there are debian init scripts included so I think Paul
> knew what he was doing ;)

3 years :-)

2009-2011 sysvinit
2011-2015 systemd

however, you just need to copy the status function from another sysvint 
scipt and modify the logic, it's not much more than read the pidfile and 
look if the process is running

attached the fedora init-script from 2010 extracted from the src.rpm at 
https://koji.fedoraproject.org/koji/buildinfo?buildID=167533, they had 
always the status function included


> 2015-10-04 18:29 GMT+02:00 Reindl Harald <[email protected]>:
>>
>>
>> Am 04.10.2015 um 18:04 schrieb Matt.:
>>>
>>> This is true for RHEL based distros, but Debian based ones are not
>>> like that yet and also think it's not wise to depend on the OS for
>>> service status, I still like the old fashioned way here.
>>
>> you are aware that in teh recent Denian systemd ist the default init-system?
>> "not wise to depend on the OS for service status" - who else than the OS? a
>> random script reading some textfile and calling wrapped external commands is
>> not really relieable for a status
>>
>>
>>> 2015-10-04 17:52 GMT+02:00 Reindl Harald <[email protected]>:
>>>>
>>>>
>>>> Am 04.10.2015 um 17:40 schrieb Matt.:
>>>>>
>>>>>
>>>>> Has anyone a working status addition to the init script for checking
>>>>> the service status per service ?
>>>>>
>>>>> There was some in the .sh scriptf for the 1.x versions but not in th
>>>>> 3.x versions.
>>>>>
>>>>> Would be nice to see some example as I'm trying something but am not
>>>>> sure
>>>>
>>>>
>>>> consider using systemd - the whole conept about status in sysvinit is
>>>> broken
>>>> by design and just a workaround - why? because sysvinit has and never had
>>>> a
>>>> solid clue what a service really does
>>>>
>>>> a init system which give relieable status back needs to monitor the
>>>> main-PID
>>>> and not rely on PID files somewhere, well and it can monitor services and
>>>> restart them automatically while a manual stop compared to monit or
>>>> similar
>>>> tools would not start the service again (and yes dbmail really needs some
>>>> monitoring, only the crashes last weekend of our imapd would have been
>>>> enough to driving an admin crazy when need to intervention each time)
>>>>
>>>> [root@srv-rhsoft:~]$ systemctl status dbmail-imapd.service
>>>> ● dbmail-imapd.service - DBMail IMAP Server
>>>>      Loaded: loaded (/usr/lib/systemd/system/dbmail-imapd.service;
>>>> enabled;
>>>> vendor preset: disabled)
>>>>      Active: active (running) since Mi 2015-09-30 12:55:27 CEST; 4 days
>>>> ago
>>>>    Main PID: 8693 (dbmail-imapd)
>>>>      CGroup: /system.slice/dbmail-imapd.service
>>>>              └─8693 /usr/sbin/dbmail-imapd -D

_______________________________________________
DBmail mailing list
[email protected]
http://mailman.fastxs.nl/cgi-bin/mailman/listinfo/dbmail

#!/bin/bash
#
# Startup script for the DBMail Imap Server
#
# chkconfig: - 81 19
# description: DBMail is a mail server with a database backend.
# processname: dbmail-imapd
# pidfile: /var/run/dbmail-imapd.pid
# config: /etc/dbmail.conf
### BEGIN INIT INFO
# Provides:          dbmail-imapd
# Required-Start:    $local_fs $network $syslog
# Should-Start:
# Required-Stop:
# Default-Stop:      0 1 2 6
# Short-Description: Start dbmail-imapd daemon
# Description:       dbmail-imapd is the imap interface to the dbmail system.
### END INIT INFO

# Source function library.
. /etc/rc.d/init.d/functions

DBLIBDIR=/usr/lib/dbmail
export LD_LIBRARY_PATH=$DBLIBDIR:$LD_LIBRARY_PATH

if [ -f /etc/sysconfig/dbmail-imapd ]; then
        . /etc/sysconfig/dbmail-imapd
fi

# Path to the dbmail script.
exe=/usr/sbin/dbmail-imapd
prog=dbmail-imapd
RETVAL=0

start() {
    echo -n $"Starting $prog: "
    daemon $exe
    RETVAL=$?
    echo
    [ $RETVAL -eq  0 ] && touch /var/lock/subsys/$prog
    return $RETVAL
}
stop() {
    echo -n $"Stopping $prog: "
    killproc $exe
    RETVAL=$?
    echo
    [ $RETVAL -eq 0 ] && rm -f /var/lock/subsys/$prog /var/run/$prog.pid
}
reload() {
    echo -n $"Reloading $prog: "
    killproc $exe -HUP
    RETVAL=$?
    echo
}

# See how we were called.
case "$1" in
  start)
    start
    ;;
  stop)
    stop
    ;;
  status)
    status $exe
    RETVAL=$?
    ;;
  restart)
    stop
    start
    RETVAL=$?
    ;;
  condrestart)
    if [ -f /var/run/$prog.pid ] ; then
        stop
        start
    fi
    RETVAL=$?
    ;;
  reload)
    reload
    ;;
  *)
    echo $"Usage: $prog {start|stop|restart|condrestart|reload|status}"
    RETVAL=3 
esac

exit $RETVAL

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iEYEARECAAYFAlYRV5gACgkQhmBjz394AnmhqQCdEXmLrpAX4b/7RQ+MONPtMp74
oxIAn1159ixWe0oHFX5b3C/RCdvSt2gQ
=czJj
-----END PGP SIGNATURE-----