[collectd] Nagios Perfdatas & Collectd

[Cyril Feraudet]

Hi Sebastian,

I need at work to have one unique centralized graph generator with a 
unique
web interface.

I was thinking about a Nagios event broker module, with a 
perdata->collectd
mapping config file, who push data in unixsock like socket. (In fact it 
would be
the jsonrpc collectd plugin of Yves Mettier :
https://github.com/perfwatcher/collectd-pw/tree/plugin-jsonrpc/5.1.1)

Regards,
Cyril Feraudet

> Hi,
>
> On Tue, Nov 20, 2012 at 11:49:24AM +0100, Cyril Feraudet wrote:
>
>> Is there anyone using Collectd to graph Nagios Perfdata ?
>
> I'm not doing that so far but I just recently had the idea to write a
> gearman worker plugin which may receive the perfdata send by 
> mod-gearman
> ;-)
>
> Anyway, I'm not sure if that makes sense after all. We would probably
> need some kind of mapping of Nagios services to collectd instances. 
> We
> could use a generic mapping like
> <nagios-hostname>/nagios-<service-desc>-<label>/<type> (i.e. similar 
> to
> PNP4Nagios with RRD_STORAGE_TYPE = MULTIPLE) and then let the user
> specify a default type and a list of exceptions for that (similar to
> DATATYPE in the PNP4Nagios check_commands configuration). Anything 
> more
> flexible would probably require quite some (user configuration) work.
>
> Cheers,
> Sebastian