Replication issues

[Lun4m]

While performing the migrations, the standby node somehow got detached from the primary node.
Unfortunately this in turn blocks ingestion to the primary, since it keeps waiting for the standby to come back up (we are using synchronous replication).
I tried running repmgr standby follow, but it complains with "standby node not found in 'pg_stat_replication'".
The "quick" workaround I found was to run the following:

repmgr standby unregister
sudo systemctl stop postgresql
repmgr standby clone -h <primary_ip> -U repmgr -d repmgr -F
sudo systemctl start postgresql
repmgr standby register

but this is not ideal since all the data needs to be copied from scratch.

Anyway, I'm not sure why the standby stopped working in the first place. Maybe we should use replication slots, modify the WAL options, or use Barman.

We should also implement a way to detect when a node goes down, be it primary or standby.