Re: [Nagios-devel] Peformance Data in a Database
Posted: Fri Apr 16, 2004 12:32 am
Sorry once more.
Hello Ben.
I think some effort must undertaken against dispersion
of configuration data over several sources.
Essentially
> metric:
> +------+---------+--------+------+
> | host | service | metric | unit |
> +------+---------+--------+------+
is part of that.
I offer you to discuss the way configuration of hole network
could be represented in more generic and structured manner
than plaintext used and place where this have to reside.
I guess it may help in making "path calculating" feature
or other that structure dependant.
Isn't this topic too academic?
----------------------------------------------
On Thu, 15 Apr 2004 16:11:28 +0100
Ben Clewett wrote:
> Chris Wilson wrote:
>
................
>
> Ideally there would be one table for all performance data. This line
> would form two records in this table. With a parent table describing
> the metric, this with a parent table of the service, as exists, and a
> parent table of host, as exists.
>
> perfdata:
> +------+---------+--------+----------+---------------------+
> | host | service | metric | value | time |
> +------+---------+--------+----------+---------------------+
> | host | http | time | 0.002143 | 2004-04-15 16:00:00 |
> | host | http | size | 1132.000 | 2004-04-15 16:00:00 |
> +------+---------+--------+----------+---------------------+
>
> metric:
> +------+---------+--------+------+
> | host | service | metric | unit |
> +------+---------+--------+------+
> | host | http | time | s |
> | host | http | size | B |
> +------+---------+--------+------+
>
> Therefore another reason for having 'hosts', as the logical
> informational domain describing a relation onto which services can be
> referenced.
>
> > Right now I can't get Nagios to log any performance data at all, so
> it may be a moot point
>
.............
Best regards.
Druzhinin Eugene.
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]
Hello Ben.
I think some effort must undertaken against dispersion
of configuration data over several sources.
Essentially
> metric:
> +------+---------+--------+------+
> | host | service | metric | unit |
> +------+---------+--------+------+
is part of that.
I offer you to discuss the way configuration of hole network
could be represented in more generic and structured manner
than plaintext used and place where this have to reside.
I guess it may help in making "path calculating" feature
or other that structure dependant.
Isn't this topic too academic?
----------------------------------------------
On Thu, 15 Apr 2004 16:11:28 +0100
Ben Clewett wrote:
> Chris Wilson wrote:
>
................
>
> Ideally there would be one table for all performance data. This line
> would form two records in this table. With a parent table describing
> the metric, this with a parent table of the service, as exists, and a
> parent table of host, as exists.
>
> perfdata:
> +------+---------+--------+----------+---------------------+
> | host | service | metric | value | time |
> +------+---------+--------+----------+---------------------+
> | host | http | time | 0.002143 | 2004-04-15 16:00:00 |
> | host | http | size | 1132.000 | 2004-04-15 16:00:00 |
> +------+---------+--------+----------+---------------------+
>
> metric:
> +------+---------+--------+------+
> | host | service | metric | unit |
> +------+---------+--------+------+
> | host | http | time | s |
> | host | http | size | B |
> +------+---------+--------+------+
>
> Therefore another reason for having 'hosts', as the logical
> informational domain describing a relation onto which services can be
> referenced.
>
> > Right now I can't get Nagios to log any performance data at all, so
> it may be a moot point
>
.............
Best regards.
Druzhinin Eugene.
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]