Measuring and Reporting Uptime
Methodology for measuring availability and reporting to customers
How availability is measured
Platform availability is monitored continuously via New Relic Ping Monitor, which checks the main endpoint and generates a Critical priority alert whenever the service becomes unavailable.
- Measurement source: periodic synthetic checks of the public endpoint
- Error detection: NRQL alerts on application error rates (an additional degradation signal)
- Runtime exceptions: captured by Sentry
Details about the monitoring stack: Metrics and Alerts.
Availability calculation
Monthly availability is calculated as:
Availability (%) = (Total minutes − Unplanned downtime minutes) / Total minutes × 100Where planned (announced) maintenance minutes are excluded from downtime — see Maintenance Windows.
Reporting
Differs by model
Shared Model (SaaS): current status and history are available on Status and History; customers can request availability reports to verify compliance with the 99.9% SLA. Enterprise Model (Dedicated): availability and SLA compliance reporting is done according to the terms set out in the contract.
Incidents
Events affecting availability are handled in accordance with the Incident Response Procedure, and affected tenants are notified by email.