aws postgresql OldestReplicationSlotLag increase continuous [DBZ-4778]

Migrated from [DBZ-4778](https://issues.redhat.com/browse/DBZ-4778)

After cdc connection using postgresql connector, OldestReplicationSlotLag continues to increase as shown below.

I think it's a wal disk space consumption issue.Please check if there is something wrong with my settings

here is my cdc config 

publication.autocreate.mode=disabled
publication.name=cdc_publication
plugin.name=pgoutput
heartbeat.interval.ms=1000
heartbeat.action.query=update public.cdc_heart_beat set id = 1, last_heart_beat = current_timestamp where id = 1;

i'm using 

confluent platform 5.5.1 on aws ecs cluster 

debezium 1.7.1

Start by reproducing the continuously increasing OldestReplicationSlotLag with the PostgreSQL connector configuration shown, including the pgoutput plugin, disabled publication autocreation, and heartbeat.action.query. Compare replication-slot and WAL behavior in the stated Debezium 1.7.1 and Confluent Platform 5.5.1 environment; done means identifying whether a setting or connector behavior causes the lag and documenting the verified resolution.

Written by the indexing model from the issue text.