Re: Backup history file should be replicated in Streaming Replication?

Tom Lane <tgl@sss.pgh.pa.us>

From: Tom Lane <tgl@sss.pgh.pa.us>
To: Fujii Masao <masao.fujii@gmail.com>
Cc: Simon Riggs <simon@2ndquadrant.com>, Heikki Linnakangas <heikki.linnakangas@enterprisedb.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2009-12-22T19:22:01Z
Lists: pgsql-hackers
Fujii Masao <masao.fujii@gmail.com> writes:
> On Wed, Dec 23, 2009 at 4:09 AM, Simon Riggs <simon@2ndquadrant.com> wrote:
>> It's needed to find the place where the backup stopped, so it defines
>> the safe stopping point. We could easily pass that info via WAL, when
>> streaming. It doesn't actually matter until we try to failover.

> Right. And, it's also needed to cooperate with HS which begins accepting
> read-only queries after a recovery reaches that safe stopping point.

OK, so the information needed in the WAL record doesn't even include
most of what is in the history file, correct?  What you're actually
talking about is identifying a WAL position.  Maybe you'd want to
provide the backup label for identification purposes, but not much else.
In that case I concur that this is a better solution than hacking up
something to pass over the history file.

			regards, tom lane