Postgresql replication slot used by barman has a long replay_lag

pid WHERE _type = 'logical'; slot_name On pg_stat_activity I see the logical replication worker of this table always active

Postgresql replication slot used by barman has a long replay_lag But when you stop instances, they don't free replication slots they use replication worker for subscription 16739 exited with

pid WHERE _type = 'logical'; slot_name On pg_stat_activity I see the logical replication worker of this table always active

replication slot does not exist --slot : It causes the WAL streaming to use the specified replication slot active column, it is active now ! postgres=> SELECT * FROM

Replication slots are used to safeguard against the premature deletion of WAL segments before they have been applied by the standby server, thus

replication slot is active for pid And this finally explains why that inactive replication slot causes the retention of that much WAL on an idle database: there actually are some

Currently postgres doesn't allow dropping a replication slot that's active This can make certain operations more time-consuming or stuck in production