Re: Important lag in replication streaming between master and standby

On Mon, Aug 26, 2019 at 5:21 PM Mai Peng <[email protected]> wrote:
>
> Hello,
>
> I’ve got several times this error : "requested WAL segment 000000020000A01A0000004F has already been remove". I think it’s due to the lag between our master in EU zone and our standby in US zone.
> How could we configure Postgres to handle this issue?

As others already stated, you can use either wal_keep_segments or
replication_slots with advantages and drawbacks on both sides. One
thing to point out, also, since you did not provide a lot of
information, is that if you have another slave consuming wals, it
could remove (archive_cleanup) segments ignoring other slaves.

Personally, I would go for a replication slot and see if the lag can
be kept under control (i.e., the pg_wal directory does not grow too
much).

Luca