Release Safety Monitoring

Release safety monitoring for React Native

Shipping is the risky step, not the last one. Ubriot watches each release for about 15 minutes after it lands, alerts on a backend regression, and offers a one-click rollback. The same check runs continuously for uptime.

What release safety watches for you

  • Post-release backend health watch (~15 minutes)
  • Alerts on down (timeout, network error, 5xx)
  • Alerts on degraded (non-2xx or slow response)
  • One-click OTA rollback to the previous update group
  • Continuous uptime checks, 24h and 7d
  • Per-app health status, HTTP code, and latency
  • Auto-detect the backend health endpoint
  • Runs on the same release pipeline as your builds and OTA

Warn, ship, watch: the full release arc

Ubriot covers the release before it goes out, as it ships, and after it lands.

Step 1

Warn before you ship

Release intelligence checks the release before it goes out. Before an OTA, Ubriot warns you if the update would downgrade the runtime your users are on, so obvious mistakes never leave the gate.

Step 2

Ship the release

Publish an OTA to a channel with ubriot update, or send a build to TestFlight. Ubriot records the release reference and the channel it went out on.

Step 3

Watch backend and launch signals

Right after the release, Ubriot polls your backend health endpoint and reads launch outcomes from ubriot-updates-rn. It records backend checks, launch counts, crashes, and crash-free percentage beside the release.

Step 4

Alert on down, degraded, or crash-heavy

If the backend times out, returns a weak response, or the app-side launch signal crosses the crash threshold inside the watch window, Ubriot marks the release watch degraded and alerts you.

Step 5

One-click rollback

When a release looks unhealthy, roll the channel back to the previous update group from the dashboard. Users get the previous bundle on their next launch.

How it knows something broke

Release safety watches two signals: your app's backend health endpoint and launch outcomes reported by ubriot-updates-rn. After a release, Ubriot polls that URL for about 15 minutes: a timeout or 5xx marks it down; a non-2xx or slow response, or a crash-heavy first-launch slice on the new OTA group, marks it degraded. Either alerts you with a one-click rollback, and the checks keep running after the window for plain uptime (24h and 7d numbers).

The honest limit

Crash signal depends on the app using ubriot-updates-rn and reporting launch outcomes. It does not claim native crash-symbolication, JavaScript stack traces, or full session replay. Use release safety for what it is good at: catching backend regressions and crash-heavy OTA launch slices early, then getting the channel back to the previous bundle fast.

Turn it on in two steps

Point Ubriot at your backend, flip the toggle, and every release from then on is watched.

Add your backend health URL

Paste your health URL, or give Ubriot your app domain and let auto-detect find it. Ubriot probes the common paths, /health, /healthz, /api/v1/health, /status, and more, then shows you which ones respond.

Toggle monitoring on

Enable "Monitor this backend" and save. Ubriot checks the URL on a schedule, alerts you when it goes down, and watches it after each release. Open release health any time to see status, uptime, and roll back.

Watch your next release, not just your builds

Add a backend health URL, toggle monitoring on, and Ubriot watches every release with a one-click rollback ready.

Get started