Skip site navigation (1) Skip section navigation (2)

Re: Replication for migration

From: Scott Marlowe <scott(dot)marlowe(at)gmail(dot)com>
To: Kenneth Marshall <ktm(at)rice(dot)edu>
Cc: Martin Spinassi <martins(dot)listz(at)gmail(dot)com>, pgsql-admin(at)postgresql(dot)org
Subject: Re: Replication for migration
Date: 2009-05-14 22:33:36
Message-ID: (view raw, whole thread or download thread mbox)
Lists: pgsql-admin
On Thu, May 14, 2009 at 7:59 AM, Kenneth Marshall <ktm(at)rice(dot)edu> wrote:
> On Thu, May 14, 2009 at 10:57:26AM -0300, Martin Spinassi wrote:
>> Hi list!
>> I've to do a migration of postgresql from version 8.2 to 8.3 in one
>> server. The problem I see here is, to avoid data loss (people still
>> using the web site that uses postgres), I'll need to stop postgres 8.2,
>> make a pg_dumpall, and then restore that information in 8.3.
>> Taking some considerations, like database size, server, etc. I estimate
>> that it will be at least 30 minutes of downtime, and that's pretty much
>> time.
>> I've thinking in the possibility of doing some kind of master/slave with
>> both versions (I must change the default port in one of them, as
>> everything must happen on a single machine), and find some useful links
>> like:
>> 2C_and_Connection_Pooling
>> I'd love to read some experiences or advices, since I haven't done it
>> before. Also if you have some link o document that you think I must
>> read, it'll be welcomed.
>> Thanks to all.
>> Mart??n
> I used Slony replication to upgrade a database from 8.2 -> 8.3 and
> it worked quite well. You will need an outage to change the master
> to the 8.3 database and re-point your apps, but if scripted, that
> is minimal.

Where I work we use slony as our standard replication platform, and we
used it to migrate from 8.1 to 8.3.  Went smooth as silk.  We had
about 30 seconds of downtime on a Saturday night, our slowest time.

In response to

pgsql-admin by date

Next:From: J. Carlos MuroDate: 2009-05-15 10:44:39
Subject: Schema best practices
Previous:From: S ArvindDate: 2009-05-14 16:41:13
Subject: Re: pg_top problem

Privacy Policy | About PostgreSQL
Copyright © 1996-2017 The PostgreSQL Global Development Group