Monitoring is decentralised. Incident Response is distributed based on timezone, availability and project knowledge.
We use a number of services to maintain effective monitoring:
- New Relic - Watches server statistics, allows us to ensure great performance and stability of our servi es
- Uptime Robot - Checks that our servers are up. Screams if they aren't.
- VictorOps - Incident Response Management. Dispatches alerts to sysadmins and compiles a nice timeline for us to manage incidents.