Hans du Plooy wrote:
> Hi guys,
> 
> I'm seeing something weird.  I maintain a fair number of mailservers at
> clients' sites.  Some are Debian Sarge, but most are some version of
> SUSE from 9.1 though 10.1 as well as SLES9.  On all of them I run
> postfix with amavisd-new and sqlgrey (http://sqlgrey.sourceforge.net/).
> sqlgrey has a simple startup script - see below - that I put
> in /etc/init.d/
> 
> For some reason, on all the SUSE machines, sqlgrey don't start at boot.
> I have used insserv to add it to the runlevel, I've made symlinks by
> hand to S99 to make sure it starts after the stuff it depends on, I've
> en put it in boot.local.  Reboot the box, and again, it's not started.
> But when I start it by hand, it starts normally.
> 
> I haven't been able to go to one of them and see what comes up on the
> screen - most of them can't be switched off during business hours.
> 
> Does anyone know why it wouldn't start?

OK, never mind, I didn't read the original message. Darix is exactly right.

messing around with symlinks is error prone and tedious, I just use the suse
mechanisms already in place, and let the system do the grunt work.

Here is the rc script I use - drop it in /etc/init.d/, make a symlink to
/usr/sbin/rcsqlgrey, and run chkconfig -a sqlgrey.

Starts every time - not that I ever reboot, but those times when power
failures have occurred, everything always comes up running as it should.

J



#!/bin/sh
#
# Copyright (c) 2000-2006 Mirai Consulting 2001-2003

# Author: Viktor Vogel <[EMAIL PROTECTED]> for SuSE 10.0

#
# sqlgrey 
#

# /etc/init.d/sqlgrey
#
### BEGIN INIT INFO
# Provides: sqlgrey
# Required-Start: 
# X-UnitedLinux-Should-Start: $ALL $portmap ipsec named dhcpd
# Required-Stop:
# X-UnitedLinux-Should-Stop: $ALL
# Default-Start: 3 4 5
# Default-Stop: 
# Short-Description: sqlgrey
# Description: sqlgrey implements greylisting service for postfix
### END INIT INFO

SQLGREY=/usr/sbin/sqlgrey
STARTOPTS=' -d'
KILLOPTS=' -k'
SGCFG=/etc/sqlgrey/sqlgrey.conf
SGPIDFILE=/var/run/sqlgrey/pidfile

. /etc/rc.status
rc_reset

# See how we were called.
case "$1" in
  start)

        echo -n "Starting sqlgrey "
        rc_status -v

        cd ~sqlgrey
        $SQLGREY $STARTOPTS

    ;;

  stop)

        echo -n "Shutting down sqlgrey "
        rc_status -v
        
        $SQLGREY $KILLOPTS


  ;;

  restart)
        rc_status -v
        $0 stop
        sleep 1
        $0 start
        ;;

  status)
        rc_status -v
        ;;

  *)
        echo "Usage: $0 {start|stop|restart}"
        exit 1
        ;;
esac

rc_exit

Reply via email to