Preview
- Availability
- Best effort
- Recovery point
- 24 hours
- Recovery time
- 1 business day
- Support
- Project contact
No contractual SLA. Suitable for evaluation and self-hosted pilots.
Service levels · Transparent by design
Commit only to what is measured and operated.
The current Revolution Road preview has no contractual SLA. Standard and high-availability targets become commitments only after their prerequisites and written agreement are active.
Standard
- Availability
- 99.9% target
- Recovery point
- ≤ 24 hours
- Recovery time
- ≤ 8 hours
- Support
- Business-hours response
Requires paid database, external monitoring, nightly backups, and quarterly restore drills.
High availability
- Availability
- 99.95% target
- Recovery point
- ≤ 5 minutes
- Recovery time
- ≤ 2 hours
- Support
- 24×7 Sev-1 paging
Requires PITR, warm reporting replica, multi-zone cache, WAF, two-region recovery plan, and an executed service agreement.
Measurement rules
One definition before the first incident.
- Availability is measured at the authenticated application readiness endpoint in five-minute windows.
- Scheduled maintenance announced at least 72 hours ahead is reported separately.
- Recovery objectives are proven by a successful restore drill, not inferred from backup completion.
- Tenant-caused outages, upstream identity outages, and force majeure are documented before any exclusion is applied.
- Incident updates follow the severity matrix and every Sev-1 receives a written post-incident review.
Activation planning