Hi @mateusz.henicz, thank you for the detailed answer. The replication slot was marked as wal_status: extended, with several bytes behind. I tried issuing CHECKPOINT commands, but nothing got Postgres to update that slot. I have finally dropped the replication slot in question, following your recommendation, but it was automatically created soon after. It has been 20 hours since I did that, and the slot is again several bytes behind, and marked as active: false and wal_status: extended.
Which brings the question: Is something wrong with my Patroni cluster configuration? The replicas are all marked as stable, and their data seems fine.
