pglogical lag due to replication timeout

Hello all.

I've recently configured logical replication on postgres 9.6.5 using
pglogical 2.1.1.

The problem I'm seeing is pretty constant stream of these errors on the
service side in the postgres logs:

LOG: terminating walsender process due to replication timeout

After so many of these, I can see on the replication target side, that
things are not being updated. Eventually, the source server seems to retry
and then all the updates go through, but it seems like there's more lag than
there really should be due to all these "replication timeouts".

I'm not seeing anything on the network side, it's really a pretty fast
connection between the 2 servers, I can't see that being the issue. Plus I
have the wal_sender_timeout set to 600 (10 minutes) which seems like it
should be way overkill.

Any advice on this?

Thanks,
Greig Wise

--
Sent from: http://www.postgresql-archive.org/PostgreSQL-general-f1843780.html