Root cause
A saved binlog position is no longer available. Retention, manual purge or a topology change can cause this; downtime duration alone does not prove it.
How to fix
- Compare the saved offsets with available binlogs and the server identity. Set retention from measured change volume and the maximum outage/recovery budget.
- when_needed can trigger a new snapshot where supported, but a snapshot cannot restore every intermediate change from purged logs.
- Deleting and recreating a connector with the same identity does not necessarily clear its stored offsets. Use the documented stopped-connector offset/reset recovery procedure only with a downstream reconciliation plan.
Recovery guidance reviewed 2026-10-27 against v3.7.0.Final.
Official Debezium documentation ↗