Reliability
Service-level objective status and incident history. Aggregate metrics only; no tenant-specific data is exposed on this page.
SLO targets and measured status (24h window)
These are our internal objectives, not a contractual SLA. A tile reading “no data” means the measurement pipeline recorded no sample for that window — it is not an assertion that the objective was met. As of 11 August 2026 no SLI has yet recorded a measured value.
Loading…
Incident history
No customer-impacting incidents on the public history yet. SEV1 / SEV2 incidents are published here after the post-incident review completes.
Our commitments
These are commitments we are making, stated alongside what is actually in place today. Where a commitment has not yet been exercised, it says so.
- Customer comms for SEV1: within 1 hour of incident declaration; this page updated within 15 minutes. No SEV1 or SEV2 incident has been declared to date, so the commitment is untested.
- Disaster recovery: targets are RTO under 2 hours and RPO in minutes, by restoring PostgreSQL to a new server from point-in-time backup (35-day window, in-region only). No recovery drill has been performed to date, so treat both targets as untested. A scripted attended drill exists; quarterly cadence begins once it has run once.
- Canary deploys: API changes ramp 0% → 10% → 50% → 100% with an error-rate and latency check against a pre-deploy baseline at each traffic-bearing stage, and automatic rollback on regression. Web changes ramp on the same schedule with a timed bake but no SLO gate.