AW: Cascading Services/Service hierarchy

[Andreas Ericsson]

Mohr James wrote:
> I'm actually thinking of the reverse. Yes, it is possible that the
> HTTP service is running but the site is down. It could be any number
> of situations, such as no access to a database. So from the SLA
> perspective, the service we are providing (the web site) is down.
> 
> However, if the HTTP service is down, the site is also not accessible
> (it cannot be!) and again from the SLA perspective, the service we
> are providing (the web site) is down. If the machine crashes. The
> machine service is down, thus the HTTP service is down, thus the Web
> site service is down. With services in HP VantagePoint Operations
> (VPO), the machine goes down and the service is red, this propogates
> to the HTTP service and then to the site service.

This is exactly what happens in Nagios as well, but the services show up 
as pink and are displayed as being "on trouble hosts" instead of bright 
red in tactical overview and friends. Nagios will also suppress alerts 
for services on hosts in down state, since it's pretty obvious that 
whatever services are provided by a host can't possibly be available 
when the host is down.

> I see that the site
> service is down and then drill down to see the root cause (the
> machine crash).
> 

Nagios saves you the trouble of drilling (actually, it leaves drilling 
as an option that isn't strictly necessary). If you set up your 
notifications to provide a link to the status-page you'd be one click 
away to see what services are affected by the host being down.

> With Nagios it seems like this kind of propogation is not possible.

Yes it is. In the host -> service perspective it's automagic, so you 
don't have to configure it.
In the service -> service perspective you need to set up a 
servicedependency.

> Although I obvuously need something that monitors to the site
> services (which we already have), the fact that the machine goes down
> does not change the status of the site service. If this machine also
> had FTP, and mail, these services would still be green as well.

No, they would be red.

>  On
> the other hand, the machine going down would turn the VPO services
> red as well (because they are also not accesible).
> 
> I am not saying the the way VPO does it is the "right way". I am just
> trying to compare how Nagios does it compared to VPO.
> 

I haven't had a look at VPO, but it sounds like Nagios and VPO works in 
a similar fashion, although Nagios lets you configure some of the 
options that VPO rams down your throat.

-- 
Andreas Ericsson                   andreas.ericsson at op5.se
OP5 AB                             www.op5.se
Lead Developer


-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
Project Admins to receive an Apple iPod Mini FREE for your judgement on
who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null