Speeding up replication startup/recovery

On Wed, Apr 13, 2011 at 10:03 AM, Henry C. <henka@cityweb.co.za> wrote:

My test replication seems to be proceeding normally, but the process appears
to be quite slow:

SLAVE
...postgres: startup process   recovering 000000010000019000000024    (1)
...postgres: wal receiver process   streaming 190/244FEA80

MASTER
...postgres: wal sender process replicator 1.1.1.1(55390) streaming 190/244FEA80

There are quite a few log files to process and both machines are not heavily
taxed.  Is there any way to expedite this initial recovery process (1)?  It
seems to be chugging along at a rather sedate pace.

The replication lag is zero since the master and slave WAL locations match.

There seems to be nothing to expedite... why do you say it is slow?

Maybe because you see this as an "initial recovery process". The
recovery process remains active while processing continues.

--
 Simon Riggs                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services

On Wed, April 13, 2011 11:16, Simon Riggs wrote:

seems to be chugging along at a rather sedate pace.

The replication lag is zero since the master and slave WAL locations match.

There seems to be nothing to expedite... why do you say it is slow?

Maybe because you see this as an "initial recovery process". The
recovery process remains active while processing continues.

I think you're right - I'm not sure what I was expecting, but I need to be a
bit more realistic, I think. The rsync took an hour or two, which means there
are a lot of updates to catch up on (recovery has been at it for several hours
now).

Cheers
Henry