MirrorMonitoring

Monitoring comes in two flavors. I recommend both.

Monitoring with your own monitoring system

I strongly urge using your own monitoring system, and notifying a team of people who is familiar with your monitoring system. They are more likely to respond to internal alerts vs external.

Configure your monitoring system:

• check https://validator.test-ipv6.com/check/SITENAME
  • this does ~100 checks, the same that the https://validator.test-ipv6.com UI does.
  • checks for DNS, HTTP headers, HTTP content, MTU, TLS, content version, and various permutations
• low frequency checking is OK - anywhere from 5 minutes to 60 minutes
• recommended "first alert" at 4 hours, with repeat alerts at 24 hours

If your mirror is important to your support staff, I'd additionally:

• a monitor for your IPv4, IPv6, and mtu1280 URLs.
• with alerting/sensitivity as appropriate for your business need

External Alerts

Officially listed mirrors, I require a contact address to send alerts.

• from nagios@test-ipv6.com
• to: [your address]
• first alert is at 4 hours - ignores transient errors.
• repeat alerts at 24 hours
• final alert on observed recovery (should be observed within 5 minutes of actual recovery)