How do you handle odd contacts?

Furnish, Trever G TGFurnish at herff-jones.com
Thu Jun 26 16:19:54 CEST 2003


Well, crap!  I feel like an idiot.  Thanks, sg and Marc.

The crux of my problem was that even though both the documentation and my
current services.cfg file clearly say "host", not "hostgroup", so the entire
time that I've been considering this it's been under the mistaken belief
that I was applying services to hostGROUPS instead of to hosts.

Thank you for your help and patience. :-)

Sincerely,
Trever

-----Original Message-----
From: Subhendu Ghosh [mailto:sghosh at sghosh.org]
Sent: Wednesday, June 25, 2003 10:01 PM
To: Furnish, Trever G
Cc: Nagios-Users List (E-mail)
Subject: RE: [Nagios-users] How do you handle odd contacts?


On Wed, 25 Jun 2003, Furnish, Trever G wrote:

> Thanks, but I think either I'm doing a poor job of communicating or I'm
> making bad assumptions...
> 
> Maybe an example will help.  I currently have:
> 
> In contactgroup unixadmin, the following contacts: corp1, corp2
> In contactgroup winadmin, the following contacts: corpA, corpB
> In hostgroup unix, the following hosts: unix1, unix2, unix3
> In hostgroup windows, the following hosts: win1, win2, win3
> In service def telnet, the following hostgroup: unix
> In service def smb, the following hostgroup: windows

add:
  In contactgroup region3admin, the following contact: r3admin

for host win3, service smb: contactgroups = winadmin,region3admin

Not because templates do not have additive inheritance, the contact_group 
members must be explicit in the service definition.


-sg

> 
> My newly interested remote site admin (region3admin) is responsible only
for
> unix3 and win3.  I think what you are saying is I can create a new,
regional
> hostgroup containing hosts unix3 and win3, and also assign to it a new
> contact group containing only newadmin.
> 
> So I create contactgroup region3 with the following contact: region3admin
> And I create hostgroup region3 with the following hosts: unix3,win3
> 
> Now how do I tell Nagios that region3admin should be notified when the smb
> service goes down on win3 or the telnet service goes down on unix3,
without
> also alerting her to problems with win1, win2, unix1, and unix2?
> 
> I can't apply a telnet service check to the region3 hostgroup because only
> one of the boxes does telnet.  Ditto for smb.  Also, those services are
> already being checked once - the resulting alerts just don't go to
> region3admin, but rather to the unixadmin and winadmin contactgroups.
> 
> The only solution I can imagine (and I'm not even sure it will work) is to
> create logical hostgroups for each region (region1unix, region2unix,
> region3unix, region1win, region2win, region3win) and assign services based
> on those groups, I've effectively made my status grid page and status
> overview page pretty ugly, since I end up with a hostgroup for almost
every
> host.
> 
> I really hope I'm just missing something.  Thanks for your time and help
so
> far.
> 
> --
> Trever
> 
> 
> -----Original Message-----
> From: Subhendu Ghosh [mailto:sghosh at sghosh.org]
> Sent: Wednesday, June 25, 2003 12:38 PM
> To: Furnish, Trever G
> Subject: RE: [Nagios-users] How do you handle odd contacts?
> 
> 
> service definitions can have multiple contact-groups...
> 
> with templating in effect, there is no inheritance, so you must overwrite 
> the contact_groups entry with the contents of the template and add the 
> local admin's group.
> 
> One way to manage this it to not put contact_group info in the templates  
> that are used for sites that have local admins.
> 
> -sg
> 
> On Wed, 25 Jun 2003, Furnish, Trever G wrote:
> 
> > But then which one gets the services?  I want her to get alerts
regarding
> > the service checks as well as the host checks, but I don't want to check
> the
> > services twice (once for her, once for corporate).
> > 
> > -----Original Message-----
> > From: Subhendu Ghosh [mailto:sghosh at sghosh.org]
> > Sent: Wednesday, June 25, 2003 11:49 AM
> > To: Furnish, Trever G
> > Cc: Nagios-Users List (E-mail)
> > Subject: Re: [Nagios-users] How do you handle odd contacts?
> > 
> > 
> > You could have the host be members of multiple hostgroups - one logical 
> > and one regional.
> > 
> > -sg
> > 
> > On Wed, 25 Jun 2003, Furnish, Trever G wrote:
> > 
> > > I'm in a quandry.  I have my hosts and services and contacts nicely
laid
> > out
> > > in a functional way (unix servers, windows servers, routers, switches,
> > dns,
> > > etc) that fits nicely with the job responsibilities of the people who
> > > normally respond to outages - but now I have one remote admin at one
> site
> > > who would like to receive alerts only when one of her systems is down.
> > > 
> > > The problem is that in order to send her alerts concerning only her
> > systems
> > > (one unix box, one windows box), I think I would have to remove her
> hosts
> > > from their logical groups and create additional groups (host and
> contact)
> > > for just those systems.  That seems like a really bad solution - all I
> > want
> > > is to allow a contact to get alerts regarding a specific device,
without
> > > screwing up the logical layout of the rest of the config file.
> > > 
> > > Am I missing something?  Is there a way to "apply a contact to a host
or
> > > service" instead of the other way around?  How do you handle this at
> your
> > > site?
> > > 
> > > It seems like the way Nagios handles the connections between contacts
> and
> > > hosts and services is only partially complete.  Specificly, I can
apply
> a
> > > contact group to a hostgroup or a service, but I can't apply a contact
> (no
> > > "group") to it.  Nor can I apply a contact or contact group to a
single
> > > host.  It's nice that grouping is available, but it's not so nice that
> > > grouping is *required*. :-(
> > > 
> > > I've been using the template files for several years now but I admit
it
> > > still gives me a headache when I think about it - if I'm missing
> something
> > > I'd love any advice.
> > > 
> > > I think the situation I have is probably similar to that in lots of
> > > companies with big branch offices.  I have a core operations team at a
> > > central head quarters office, but I also have branch offices that
> usually
> > > have one or more local admins.  The local admins should only get
alerts
> > > related to their devices (in other words, based on geography); the
> > > headquarters operations team should get alerts based on their area of
> > > expertise (unix, windows, routers, etc).
> > > 
> > > The only way I can see to satisfy my requirements is to put every host
> > into
> > > its own hostgroup, just because I can't apply a single contact to a
> single
> > > host or service.  What am I missing?
> > > 
> > > --
> > > Trever
> > > 
> > > 
> > > 
> > > -------------------------------------------------------
> > > This SF.Net email is sponsored by: INetU
> > > Attention Web Developers & Consultants: Become An INetU Hosting
Partner.
> > > Refer Dedicated Servers. We Manage Them. You Get 10% Monthly
Commission!
> > > INetU Dedicated Managed Hosting http://www.inetu.net/partner/index.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
> > > 
> > 
> > 
> 
> 

-- 



-------------------------------------------------------
This SF.Net email is sponsored by: INetU
Attention Web Developers & Consultants: Become An INetU Hosting Partner.
Refer Dedicated Servers. We Manage Them. You Get 10% Monthly Commission!
INetU Dedicated Managed Hosting http://www.inetu.net/partner/index.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





More information about the Users mailing list