On Monday 13 June 2016 at 16:43:51, Tobias Köck wrote:
> We have two hosts which do the same and are a high availability solution
> for a service.
How is the high availability achieved? Are both hosts running the service at
the same time?
Can you access each host independently from outside the HA "cluster", as well
as the single HA service provided by the cluster?
> I would like to send an email if one host and/or service fails and another
> email (to an SMS provider) if both hosts/services fail (notification
> escalation).
Define three hosts.
1 - Real machine A
2 - Real machine B
3 - Virtual HA machine (A or B, depending on whichever is live at the time -
whatever it is your standard clients connect to, without realising it's an HA
solution)
Then if host 1 or host 2 goes down, you send your first email.
If host 3 goes down (ie: the entire HA cluster) you send your second email.
If the machines are both live and running whatever service it is under normal
circumstances (ie: the HA does not rely on starting the service on the
"backup" machine if the primary goes down - the service is already running and
HA just relies on switching connections to that server), then you can define
three services as well as three hosts, as above.
If the service only runs on one machine at a time, and HA relies on starting
it on the backup server when it goes down on the primary server, then you
can't monitor the service, because there's no service to be monitored on one
of the machines, but monitoring the hosts should still work.
Hope that helps,
Antony.
--
Is it venison for dinner again? Oh deer.
Please reply to the list;
please *don't* CC me.
_______________________________________________
icinga-users mailing list
[email protected]
https://lists.icinga.org/mailman/listinfo/icinga-users
