Subject: Re: Sync Rep Design
From: Simon Riggs <simon@2ndQuadrant.com>
To: Robert Treat <rob@xzilla.net>
Cc: Stefan Kaltenbrunner <stefan@kaltenbrunner.cc>,
	pgsql-hackers@postgresql.org
In-Reply-To: <AANLkTi=dr9TuYRLMWjPkJdL1Jjg+TAJXPJQbLGiN1Pag@mail.gmail.com>
References: <1293729972.1892.23804.camel@ebony>
	<4D1CC47E.7050207@kaltenbrunner.cc> <1293735853.1892.24886.camel@ebony>
	<AANLkTi=dr9TuYRLMWjPkJdL1Jjg+TAJXPJQbLGiN1Pag@mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
Date: Thu, 30 Dec 2010 20:39:05 +0000
Message-ID: <1293741545.1892.25943.camel@ebony>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit

On Thu, 2010-12-30 at 15:07 -0500, Robert Treat wrote:
> > We use a single parameter to enable synchronous replication, set in
> > postgresql.conf on both primary and standby servers:
> >
> > synchronous_replication = off (default) | on
> >
> > On the primary, synchronous_replication can be set for particular
> users
> > or databases, or dynamically by applications programs.
> >
> >
> This seems like a potential issue, where I start a server with this
> off, and then I start turning it on for specific transactions; it
> isn't exactly clear what happens, since there may or may not be a
> running synchronous rep slave available.  (I love the idea though)

Not really an issue. Even if there was a standby there a moment ago, the
standby can go away at any time. So we must cope gracefully with what
happens if you do this. By default, the parameters specify that in the
case you mention we will just use async replication (no wait!).
Options exist to change that, since some people want to wait until the
sysadmin adds a standby.

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