Multi-tier failover

[Demetri Mouratis]

On Fri, 30 Jul 2004, Jason Martin wrote:

> Particularly I'm wondering how to handle the OCSP dynamically
> determining which central server is configured to handle the
> result.

I can think of three ways to handle this:

1.	Use a load-balancer between your Distributeds and Centrals
2.	Use DNS failover to switch between your Centrals.
3.	Always update both Centrals.

In scenario 1, you have a load-balancer set up a VIP to port 5667 on
Central 1, if it is up and listening, otherwise it points to Central 2.
Most load balancers have a concept of routing weight, you could set
Central 1 to a very high weight, and Central 2 to a weight of 1.

In scenario 2, you would configure OSCP to send its updates to a CNAME
maintained in your corporate DNS, i.e. central.yourcompany.com.  In normal
circumstanes, the CNAME would point to central1.yourcompany.com.  On a
Nagios failover, part of your event handler would be to update the
corporate DNS using BIND nsupdate or some home grown approach manipulating
the zone files directly.

Scenario 3 is a fallback, not really a good design but will work if you
can't manage 1 or 2 above.  In this scenario, you setup your oscp command
to send two send_nsca commands, one each to Central 1 and Central 2. I
guess it has the advantage of easier implementation as well.

Hope that helps.
---------------------------------------------------------------------
Demetri Mouratis
dmourati at linfactory.com



-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
_______________________________________________
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