Using SeedQL

Schedules

Capture snapshots automatically on a cadence. Runs on our servers, not in your browser.

A schedule captures a snapshot of a project at a fixed interval, without anyone doing anything.

Schedules in the sidebar. Available on Business and Enterprise.

Creating one

Pick a project, give the schedule a name, choose a cadence. Presets cover every 15 minutes through daily; Custom takes a number of minutes. Minimum is 15 minutes, maximum is 30 days.

It runs on our servers

Worth stating plainly, because an earlier version of SeedQL did not: schedules fire on our infrastructure. Close the tab, shut your laptop, go on holiday. They still run.

Run history

Each schedule shows its recent runs with one of three outcomes.

Status Meaning
OK A snapshot was captured
Skipped Nothing to capture, or your plan no longer includes scheduling
Error The capture failed, with the reason

The last 20 runs per schedule are kept.

Skipped is not a failure. The usual cause is a project with no schema captured yet: there is nothing to snapshot, so the run records why and moves on rather than erroring.

Behaviour worth knowing

Run now fires immediately without moving the scheduled time. An ad-hoc capture should not postpone tonight's run.

Pausing keeps the schedule and stops the fires. Re-enabling starts a fresh interval from now rather than firing instantly because the old time has passed.

Changing the cadence re-anchors from now. Shortening from daily to hourly takes effect within the hour, not tomorrow.

After downtime, a schedule fires once. If we were down for ten hours and your schedule runs every fifteen minutes, you get one capture on recovery, not forty. Nobody wants forty snapshots.

If your plan lapses, fires stop and record as skipped. The schedule itself survives, so upgrading resumes it. Your configuration is not deleted because a payment slipped.

Monitoring

To check across every schedule at once, rather than opening each, the API exposes an org-wide run list. See Automation.