# SLA FAQ

:::note[Summary]
This page is not the contract. The [SLA](https://ohauthcat.com/legal/sla), in force from 1 November 2026, and your order form decide what applies to you.
:::

## What does the SLA cover?

Sign-in (the SSO broker) and SCIM directory sync, measured by our own external probes. The commitment is 100% availability on the Business plan and above, backed by service credits. It is a credit commitment, not a claim that the service can never go down.

## Which plans include it?

| Plan       | Uptime SLA                                                                                  |
| ---------- | ------------------------------------------------------------------------------------------- |
| Free       | None                                                                                        |
| Pro        | None                                                                                        |
| Business   | 100%, with 1× credits                                                                       |
| Enterprise | 100%, with 10× (Standard) or 25× (Premium) credits once multi-region is generally available |

## How are credits calculated?

Credit = downtime minutes ÷ minutes in the month × monthly fee × your plan's multiplier, capped at the monthly fee. Claim within 30 days after the end of the month. Credits are a discount on your next renewal, or an extension of your paid term; they are never paid in cash.

## What is excluded?

Outages of AWS or Cloudflare, your customer's identity provider, your own application or network, incorrect configuration, your use of the kill switch, announced maintenance, features marked beta, and events outside our reasonable control. The SLA lists every exclusion.

## Where does the service run today?

In one AWS region, `ap-south-1` (Mumbai), spread across its three availability zones. Losing one zone is handled automatically. The service does not yet run in more than one region; multi-region is planned and is not claimed until it is live. See [data location and backups](/trust/architecture-and-availability/).

## What happens if the whole region goes down?

At this stage, the target recovery is:

| Scenario                      | Data loss target | Time to recover target        |
| ----------------------------- | ---------------- | ----------------------------- |
| One availability zone         | None             | Automatic, seconds to minutes |
| The whole `ap-south-1` region | Up to 15 minutes | 4 to 8 hours                  |

Recovering from a region loss means rebuilding from the backup copies in `ap-south-2` (Hyderabad) and pointing traffic at it. Sign-ins and SCIM are down during that time. Once the service runs in more than one region, the target drops to seconds for sign-ins.

## What if Cloudflare goes down?

A standby route on `oauthcat.com` is designed to serve sign-ins directly from AWS during a Cloudflare outage. Cloudflare outages are excluded from SLA downtime. See [standby route behavior](/trust/standby/) for what is built and what you register ahead of time.

## Where do I see incidents?

On the [status page](/trust/status/), and through the `edge.failover.*` events described in [standby route behavior](/trust/standby/).
