Re: How to prevent master server crash if hot standby stops

On Mon, 2020-04-06 at 11:47 +0300, Andrus wrote:
> > That's why you use monitoring.
> > Never use replication slots without monitoring replication (or at least the
> > disk space on the primary).
>
> How to implement this automatically, without human interaction required ?
>
> "superuser_reserved_connections" setting exists.
> How about implementing "data_reserved_space" setting ?
>
> How to create procedure in server or maybe cron scipt which stops replication if disk becomes nearly full ?
>
> How to force to stop replication slot in master if it is in use ?

You don't.

You reserve enough disk space that WAL can grow for a while before
filling the disk. Any you need to trigger human intervention with
monitoring.

If you prefer replication to fail silently, don't use replication
slots. Use "wal_keep_segments" instead.

Yours,
Laurenz Albe
--
Cybertec | https://www.cybertec-postgresql.com