Monitoring Chrony in Grafana via Prometheus

[MaffooClock]

This is a bit off topic since it's not really specific to SatPulse, but for anyone using Chrony with SatPulse, here's a good monitoring solution for Prometheus with Grafana:

1. Prometheus Chrony Exporter by Ben Kochie (@SuperQ).
2. chrony Dashboard by equinox0815

Here's the full command to populate all charts* in the dashboard linked above (might not actually need all of the flags):

sudo chrony_exporter \
    --chrony.address=unix:///var/run/chrony/chronyd.sock \
    --collector.sources.with-ntpdata \
    --collector.tracking \
    --collector.sources \
    --collector.sourcestats \
    --collector.serverstats \
    --collector.dns-lookups \
    --collector.chmod-socket

You can also run this as a Systemd service:

[Unit]
Description=Prometheus exporter for Chrony
After=chrony.service

[Service]
Restart=on-failure
EnvironmentFile=/etc/default/chrony_exporter
User=_chrony
ExecStart=/opt/chrony_exporter/chrony_exporter --chrony.address=unix://${CHRONY_SOCK} $FLAGS
ExecReload=/bin/kill -HUP $MAINPID
TimeoutStopSec=20s
SendSIGKILL=no

[Install]
WantedBy=multi-user.target

...with the environment file at /etc/default/chrony_exporter:

CHRONY_SOCK="/var/run/chrony/chronyd.sock"
FLAGS="--collector.sources.with-ntpdata --collector.tracking --collector.sources --collector.sourcestats --collector.serverstats --collector.dns-lookups"

*Note: the "Maximum Clock Error" chart only works if you setup a recording rule as mentioned in the README.

[SuperQ]

Happy to have PRs to the exporter to improve the dashboard. As well as the Ansible role.