On 02/04/2013 01:30 PM, Simon Oosthoek wrote:
> Hi
>
> just wanted to expand on this thread...
>
> In discussions with my colleague (who's background is cacti) he insisted
> that the choice for host_name as unique identifier is a design-flaw.
> Considering that icinga2 is being built up from scratch (and
> experience), I'm wondering if it will be possible to change the host's
> name while retaining the service history if it is essentially the same
> host with a different name?
>
If I look at the config structure examples I find for icinga2, I see the
host definition looks like
object host "hostname" {
stuff
}
So I infer that hostname is still the identifier for the host.
Thinking about it some more, I suppose that although it may not be the
most canonical and pretty solution, it will usually not matter that much
in practical applications, except for the following scenario:
an understaffed noc whacks up a nagios install after half-reading the
docs. A few years later, all the staff is gone and a new understaffed
noc takes its place. After more fully, yet not completely, reading the
docs, thinks about ways to clean up the mess and cleans up the config,
on the fly also normalising the hostname conventions (there were 4
different ways of naming hosts).
A week or two later, someone asks what happened to the history of a
certain host....
I hope for all of our sakes that this is not a frequently occurring
scenario ;-)
However, assuming this doesn't just happen to me, it would have helped
me to have clearer separation between identifier and name for the host....
Cheers
Simon
------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_d2d_jan
_______________________________________________
icinga-users mailing list
icinga-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/icinga-users
