What happens when a scheduled export fails

Last updated August 14, 2026 · By the SaturnSQL team

Automatic retries

When a scheduled export fails, SaturnSQL retries it automatically rather than waiting for the next scheduled cadence. The first retry happens about 15 minutes after the failure. If it fails again, the gap between retries grows, spacing out to at most every 6 hours.

Retries keep going for 7 days from the first failure in a streak. If nothing has succeeded by then, the schedule stops running until someone restarts it. A single successful run resets everything, clearing the failure streak and putting the schedule back on its normal cadence.

Where you see the error

• The banner in the editor — yellow while the schedule is still retrying, red once it has stopped.
• The schedule icon in the saved query list — flags any query whose schedule is failing.
• The run history inside the schedule dialog, which shows every attempt and its error.

Common causes and fixes

• Query timeout — a scheduled run has a fixed time budget. See row and time limits for the current budget and how to fit a slow query inside it.
• Changed database credentials or firewall — a rotated password or a new firewall rule can block the connection SaturnSQL uses. See connecting a database to reconnect.
• Dropped tables or columns — if the query references something that no longer exists, update the SQL to match the current schema.
• Revoked Google Sheets or Slack access — reconnect the destination. See exporting to Google Sheets or exporting to Slack for the reconnect steps.

Restarting a stopped export

Once a schedule has stopped, fixing the underlying problem is not enough by itself, you also need to restart it. Any of these works:

• Retry now in the editor banner.
• Run now in the schedule dialog.
• Re-saving the schedule, even with no changes.

Your team admin can also restart it for you. Any of these clears the failure streak, and if the run succeeds the schedule goes straight back to its normal cadence.