[Wishlist] Configurable metric thresholds for nagios check_ntpmon.py #14
Description
In some use cases, the hard-coded thresholds in alert.py _metricdefs lead to unactionable critical alerts when using check_ntpmon.py.
As an example, an all-reach-mean of 20% during an edge cloud upstream BGP storm causing intermittent NTP server access may be considered by that cloud's operators as a warning indicator of upstream connectivity, but is not actionable as a critical failure of the NTP component itself.
It would be useful to have a CLI argument to provide overrides to the _metricdefs in the NTPAlerter for the nagios check.
An example simple cli implementation might look like this if one wanted to override the reach and offset thresholds:
check_ntpmon.py --check peers sync reach offset --metric-overrides 'reach:high:50:10;offset:mid:-0.1:-0.05:0.05:0.1'