[PATCH] - Schedule (host + all services )downtime via extinfo.cgi

Ton Voon ton.voon at opsera.com
Fri Sep 11 16:35:11 CEST 2009


On 11 Sep 2009, at 14:35, Thomas Guyot-Sionnest wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Ton Voon wrote:
>> In Opsview, if someone sets a downtime for the host, we use this
>> extcommand to set all the services in downtime too. We can't think  
>> of a
>> valid reason to not do that, so I agree that it should be available.
>
> I disagree, I can't think of a valid reason to do that (besides
> reporting with I don't use at that level). Since host downtime  
> suppress
> service notifications, why bother with services?

In Opsview, we found it much easier to show downtimes for services if  
the services are in a downtime state, rather than trying to calculate  
if the appropriate host is also in a downtime state.

>
> I'd be *really* annoyed if after a maintenance that took less time  
> than
> expected, of after fixing up a server that went down unexpectedly, I'd
> have to remove the downtime not only for the host but for every single
> services on it. I'd be even more if it happened on one of these  
> special
> hosts that have 30+ services associated to it. With one of the new
> Nagios instances I'm currently setting up it gets even worse with  
> hosts
> with over 60 services on it and a special one with well over 100!

This sounds like a cancel limitation - cancel downtime for host and  
for all the services too (with the same comment?).

>> I'll then get it applied to CVS.
>
> I'm not at all against this, my concerns were mainly about using this
> patch as a workaround for a reporting bug.

The patch adds an extra option to the CGIs for the behaviour that  
Hendrik is looking for. The CGIs work with the philosophy of "show all  
the available options", so it makes sense to add this.

> I would recommend making sure the CGI documentation clearly say that
> host downtime already suppress service notifications.

That's fair comment. The patch already says: "During the specified  
downtime, Nagios will not send notifications out about the host". I  
guess I can add the sentence: "Normally, a host in downtime will not  
send alerts about any services in a failed state".

My guess is that users will start to use this option in preference to  
just scheduling downtime on the host because then in the status views  
for the host, the services will show with a downtime icon. Which is  
exactly why Hendrik wants this added in.

Ton



------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july




More information about the Developers mailing list