The uptime we commit to. And what you get if we miss.

We measure availability as the percentage of 1-minute windows in a calendar month in which at least one of the three synthetic probes we run against /health returns a 2xx response from an unprivileged caller. We publish the measurement source to customer-success on request.

• Scheduled maintenance announced ≥7 days in advance.
• Emergency security patches, announced at the time.
• Force majeure events (per main agreement).
• Customer-caused incidents (e.g. exceeding a plan limit, using an unsupported API version, misconfigured SSO).
• Outages of upstream providers where we have no reasonable mitigation and have already implemented documented fallback measures.

If monthly uptime falls below the applicable threshold, you may claim a service credit equal to the percentage in the table, applied to the next invoice. Submit a claim within 30 days of month-end to sla@sutrace.io with your workspace ID and the affected window. Credits are your sole and exclusive remedy under this SLA.

• Dashboard first-byte (p95): < 800 ms at edge.
• Ingest to visible on dashboard (p95): < 10 s for metrics, < 30 s for traces/logs.
• Alert fire to first notification (p95): < 30 s.

These are operational goals, not contractual commitments.

Public at status.sutrace.io(hosted off our primary infrastructure so it stays up when we're down — see what happened to Cloudflare 18 Nov 2025). Subscribe by email or webhook from the status page itself.

For P1 (service-affecting for >10% of customers) we publish an initial update within 15 minutes, hourly updates until resolution, and a root-cause analysis within 5 business days. For regulated-industry customers needing NIS2 or DORA-shaped incident reporting, see the DPA.