Skip to main content
On every paid plan (Private, Lite, Pro and Enterprise), Meridian commits to keeping each app’s production environment available at least 99.9% of every calendar month. That leaves about 43 minutes of downtime a month. The SLA is per app: each app is measured on its own. An Enterprise agreement can set different terms.

What counts as downtime

A minute counts as down when the app’s production environment cannot answer requests because of Meridian or the infrastructure Meridian runs it on (Google Cloud). For example: the hosting platform does not route requests to the app, or the app’s managed database is unreachable. Monthly uptime is the share of minutes in the month that were not down.

What doesn’t count

  • Your own code: the app crashes, answers errors, or fails its health check after a deploy. A release that fails its health check never takes traffic, so the previous release keeps serving.
  • Limits of your plan: requests refused or slowed because the app is at its instance, database or bandwidth limit.
  • Announced maintenance, when Meridian has told you about it in advance.
  • Shopify: an outage of Shopify, its APIs or its webhook delivery.
  • Development, staging and PR previews. Only production is covered.
  • Anything you control: your custom domain’s DNS, third-party services your app calls, and changes you or your team make to the app’s settings.
  • A suspended app, for example over an unpaid invoice.
Last modified on November 13, 2026