Skip to main content
  • Outcome: you know what runs on its own, and the two screens that prove it.
  • Time: a read, about 5 minutes.
  • Before you start: Live for New Orders is behind you; if you switched platforms, Migrate too.
  • You’ll touch: the OpenSubs admin.
Every schedule below lives inside your own Supabase project, no vendor in the loop. Some work runs in your own app; if that app is down, those queues wait and drain when it’s back. Every run is recorded. A missed hour catches up on the next one; each due charge is picked up exactly once, so nothing double-charges. Your two screens:
  • Home shows a Needs attention banner only when something needs you: failed payments, or an open alert.
  • Settings > Operational health shows alerts, connected providers checked live against their own servers, scheduled jobs, webhook delivery, and queue health. Alert checks re-run every 15 minutes; the panels read live when you open the page.
Most blips fix themselves: failed payments follow a recovery schedule you control, Shopify redelivers webhooks that didn’t land, and queued work retries until it clears. When an alert opens matches each alert to its fix. The group is organized by feature, each setup guide beside its what-went-wrong pages. The morning check is the whole job most days, with When an alert opens directly beneath it. The one recurring chore, updates, closes the group.
The schedule: every minute the queues drain, every 15 minutes the watchdog checks, every hour billing and retries sweep, overnight reminders and snapshots run. All of it reports to Home and Operational health.

Four schedules, two screens. The machine works; the screens prove it did.

Last modified on November 5, 2026