[ale] Restarting if bind dies

[pwd at mdtsoft.com]

I use a package call checkservices (see rpmfind or freshmeat)
which will check on a number of services (on a number of servers) and
will run a script to re-start in which I do:

/etc/rc.d/init/SERVICE stop
/etc/rc.d/init/SERVICE start
sleep 5
/etc/rc.d/init/SERVICE stop
/etc/rc.d/init/SERVICE start


The second stop and start after 5 seconds will give a OK for both
the stop and start if all is well.

checkservices will e-mail you (I send to my cell phone as well) when
it has to re-start a service.


On  8 Feb, Jonathan Rickman wrote:
> On Thu, 8 Feb 2001, Dan Newcombe wrote:
> 
>> I am running bind 9.latest, and every so often it dies of an assertion
>> error.  I was planning on going back to 8.latest for a while, then I
>> thought   Instead of running just named, why not run a script like below.
>> 
>> #!/bin/sh
>> while (/bin/true)
>> do
>>  /usr/local/sbin/named -f
>>  #something to alert me that bind puked
>> done
>> 
>> comments?  anyone else doing something like this?
> 
> Cheops has this type of monitoring built in, if you are running X or have
> something on the same network running X.
> 
> Get it at freshmeat
> 

-- 
Philip W. Dalrymple III <pwd at mdtsoft.com>
MDT Software - The Change Management Company
+1 678 297 1001
Fax +1 678 297 1003


--
To unsubscribe: mail majordomo at ale.org with "unsubscribe ale" in message body.