Distributed monitoring without direct network connection

[Nick Lunt]

> -----Original Message-----
> From: Andreas Ericsson [mailto:ae at op5.se]
> Sent: 29 November 2008 14:01
> To: Nick Lunt
> Cc: nagios-users at lists.sourceforge.net
> Subject: Re: [Nagios-users] Distributed monitoring without direct
> network connection
> 
> Nick Lunt wrote:
> > Hi folks
> >
> > nagios 3.0.5 on RHEL 4u6.
> >
> > We have nagios servers all over the uk and we want to get all alerts
> > from each nagios server to a central nagios server at our main
> offices.
> >
> > We do not have permanent network connectivity to the remote nagios
> > servers so using NSCA is not an option.
> >
> > Has anyone any idea of how to overcome this problem ?
> 
> Queue the events that were unsendable and send them when it becomes
> possible. Merlin is designed to handle frequently failing links with
> sometimes extremely long downtimes (it already does this), but it's
> not really production level stable yet, so I wouldn't recommend using
> it for this (unless you're interested in completing it yourself or
> sponsoring me or op5 to do it for you, ofcourse).
> 
> More about merlin at http://git.op5.org/git/nagios/merlin.git
> 
> pnsca, another module available there, can probably be trivially
> rewritten to stash alerts and whatnot with very good performance.
> 
> > I am thinking of
> > getting the remote nagios servers to send email alerts to an account
> on
> > the central nagios server then trying to get an alert generated
based
> on
> > the contents of the email, has anyone tried this before ?
> >
> > Or does anyone have any better ideas for solving this problem ?
> >
> 
> That depends on what your end-goal is, really. Do you want only one
> server
> to send notifications, or do you want your central server to be able
to
> generate reports from the data sent in from the slave systems?
> 
> If only one server should send notifications, I'd recommend using a
> solution
> with lower latency that gathering everything and shipping it as an
> email.
> One-way UDP communication would be one solution here, I guess, but it
> does
> require the network to be physically present at all times (and there's
> no
> failure detection what so ever, as UDP is a fire-and-forget protocol).
> Merlin would help in this case (although it can't send over UDP yet).
> 
> If it's for reporting reasons, you'd be better off sending the
logfiles
> as
> emails when they're being rotated and then merging them together on
the
> master server. That means you can't get *accurate* reports more often
> than
> the logs are rotated, but since you'll need to sort-merge them
anyways,
> that's still going to be a problem.
> Neither merlin nor NSCA can help here, I'm afraid, as entries in the
> logs
> would get completely jumbled unless you sort-merge them before taking
> generating reports from them.

Thanks for the detailed info Andreas. I still think the "nagios event ->
email -> nagios server" is the only realistic solution. It's not perfect
as mail servers can fail and mail can get delayed but it's the best we
can do at the moment.

Cheers,
Nick. 

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
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