Soft-deprecation of prometheus/nagios_plugins

[Bjoern Rabenstein]

If you don't use or need https://github.com/prometheus/nagios_plugins,
you can stop reading now.

For everyone else, please have a look at
https://github.com/magenta-aps/check_prometheus_metric . It is a
backwards-compatible fork/rewrite of the original Nagios plugin. If it
works for you as expected, please consider migrating. If anything
breaks unexpectedly, please report it in the new repo. At some point,
I'd like to simply redirect from the old repo to the new one.

For those who like more context, here it is:

prometheus/nagios_plugins is a very very simple Nagios plugin. We
needed it back in the early days of Prometheus at SoundCloud, because
we still used Nagios for most of our alerts, and that new Prometheus
thing had to integrate with it somehow. Alertmanager was not yet ready
for production, so firing Nagios alert based on Prometheus queries
(very much in the same way as we also fired Nagios alerts based on
Graphite queries, which were based on data from statsd) kind of made
sense. Back then, there was no huge Prometheus ecosystem, so anything
we did that was remotely Prometheus related, ended up in the now
famous Prometheus GitHub org. Fast forward a few years, SoundCloud
switched of their last Nagios (or Icinga, to be precise) server. I
kept maintaining the prometheus/nagios_plugins repo, because quite a
few others found it useful for various reasons. However, by now all
but the most common integrations with Prometheus lived outside of the
Prometheus GitHub org, and the only reason why the rather simplistic
Nagios plugin had such a privileged location was historical. One day,
Emil 'Skeen' Madsen (@Skeen on GitHub) came along and proposed some
more involved feature additions. I felt unqualified to properly review
them, without any Nagios or Icinga server near me. We discussed
various options to make it still happen, and the most organic was in
the end to let @Skeen maintain a fork of prometheus/nagios_plugins in
a GitHub org of his own (in this case, magenta-aps), similar to the
many other 3rd party integrations out there. prometheus/nagios_plugins
now contains a note to consider magenta-aps/check_prometheus_metric
instead, and once the latter has proven its maturity, we might just
turn prometheus/nagios_plugins into a redirect to
magenta-aps/check_prometheus_metric.

--
Björn Rabenstein
[PGP-ID] 0x851C3DA17D748D03
[email] bjo...@rabenste.in