# /etc/init.d/sec start && /etc/init.d/sec status ; cat /run/sec.pid && ps aux|grep sec * Starting sec ... [ ok ] * status: crashed 1353 root 1353 0.0 1.1 25568 11580 ? S 10:47 0:00 /usr/bin/perl -w /usr/bin/sec -pid=/run/sec.pid -detach -log=/var/log/sec.log -conf=/etc/sec.conf -input=/var/log/messages -input=/var/db/rbldnsd/log -debug=4 After changing init script: - start-stop-daemon --start --quiet --exec /usr/bin/sec \ + start-stop-daemon --start --quiet --interpreted --exec /usr/bin/sec \ It works as it should: # /etc/init.d/sec start && /etc/init.d/sec status ; cat /run/sec.pid && ps aux|grep sec * Starting sec ... [ ok ] * status: started 3523 root 3523 0.0 1.1 25504 11720 ? S 10:50 0:00 /usr/bin/perl -w /usr/bin/sec -pid=/run/sec.pid -detach -log=/var/log/sec.log -conf=/etc/sec.conf -input=/var/log/messages -input=/var/db/rbldnsd/log -debug=4
Another option would be to use the PID file for status checking instead of the changing process name, i.e. you only have to set > pidfile="/run/sec.pid" in "/etc/init.d/sec", which should also fix the problem and is more reliable.
Fixed everywhere, but particularly in 2.7.10. Thanks for the patch.
It works, thanks.