These simple tools are used to proxy alerts generated in Prometheus AlertManager into passive alerts in Nagios.
This tool creates a set of passive nagios monitors based on potential
alerts from Prometheus. All these services will inherit from
promservice which should be defined in a different Nagios
configuration file to hold all common settings.
To have a monitor generated, the alert in Prometheus must have an
annotation named hostmap. This can either be the name of a metric,
in case all hosts with that metric will be included, or it can be a
label-filtered metric in which case only those with that label will be
included. Once the metrics are found, the hostname is extracted from
the label name.
To exclude specific servers from having monitors generated, create another
annotation named excludehosts. This will contain a comma separated
list of hosts to exclude from generation.
If the parameter --refreshstate is given, commands to refresh the
state of all alerts that are not firing will be written to this
file at the end of the run, to ensure they are not flagged as stale
in Nagios.
This script should be run as a daemon, and under a user account with write permissions on the nagios control file. It listens on a specific http port (localhost only) for AlertManager to make a webhook post to. Once it receives such a post, it will translate it into a passive command result in Nagios and write that to the nagios control file.
It also exposes an endpoint at /ping which should be monitored
actively from Nagios. It will return 200 OK when it's
working. This endpoing will also start firing if any alerts are lost
because of an exception during the processing (such as permissions
errors or misconfiguration). To reset such a state, restart the proxy
(which has no other state, so it's safe to restart).