These settings allow for a system restart without triggering the alerts
Period:
5 minutes
The expected time between pings
Grace Time:
10 minutes
When a check is late, how long to wait to send an alert
Bash script
Used to monitor if Prometheus and Alertmanager are running, and if the entire machine is offline or unresponsive.
sudo vim /etc/default/healthchecks.sh
Edit UNIQUE_ID
#!/bin/bash
# Healthchecks.io URL
# ***********
# CHANGE HERE
# ↓↓↓↓↓↓↓↓↓↓↓
healthchecks_io_url="https://hc-ping.com/<UNIQUE_ID>"
# URLs for the health endpoints of Prometheus and Alertmanager
prometheus_health_url="http://localhost:9090/-/healthy"
alertmanager_health_url="http://localhost:9093/-/healthy"
# Check health of Prometheus
if curl -f ${prometheus_health_url}; then
# Check health of Alertmanager
if curl -f ${alertmanager_health_url}; then
# Send heartbeat to Healthchecks.io
curl -fsS --retry 3 ${healthchecks_io_url} > /dev/null
else
echo "Alertmanager is not healthy."
fi
else
echo "Prometheus is not healthy."
fi
