[monit] Assistance with configuration

[Jack S]

Hello All,

OK I have read some of the documentation, and still am unclear on the
best way to handle this.
I will use my DNS monitoring as my sample:
#-----< Bind >-----
check process named with pidfile /var/named/chroot/var/run/named/named.pid
   alert address@hidden only on { timeout, nonexist }
   start program = "/etc/rc.d/init.d/named start"
   stop program = "/etc/rc.d/init.d/named stop"
   if failed host 127.0.0.1 port 53 type tcp protocol dns then alert
   alert address@hidden only on { timeout, nonexist }
   if failed host 127.0.0.1 port 53 type udp protocol dns then alert
   alert address@hidden only on { timeout, nonexist }
   if 5 restarts within 5 cycles then timeout


I'm not sure the best/right way to make this work so here is what I'm
trying to do:
1. I want to test if bind is running of course, since bind may reboot
through scripts when we update information I really don't need to know
if a pid changes ( way too many notifications that way) so I add the
only on timeout, nonexist.  I am testing 3 things, the process is
running, the 2 ports are responding which is good am I missing any
other test I should be doing?
2. I want to have it evaluate that it is actually down for a minute or
some number which I can control, what is the best way to do this? So
if it fails for 1 min, it's really down, 20 seconds it may just be
busy so don't go trying to reboot processes and such.
3.In this example will it only restart on a process issue? what if
here is a port issue, in the above example, does start and stop
execute no matter what the failure is?

I have the above configured, if I restart DNS I get notifications:
Date:        Tue, 15 Dec 2009 01:54:25 -0500
Action:      Alert
Host:        myserver.net
Description: process PID changed to 15258

PID changed Service named

Once I can figure out the logic / formatting of this language per say
I'm hoping to clean up all my notifications and utilize monit a little
better.

-- 
Thanks!
Jack