Re: clustering failover... ala Oracle Parallel server

From: "Alexander Staubo" <alex(at)purefiction(dot)net>
To: hanasaki <hanasaki(at)hanaden(dot)com>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: clustering failover... ala Oracle Parallel server
Date: 2007-08-07 16:37:11
Message-ID: 88daf38c0708070937u4596f816xb267cad53ec4d95d@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On 8/4/07, hanasaki <hanasaki(at)hanaden(dot)com> wrote:
> clustering failover... ala Oracle Parallel server

Note that OPS is now called RAC (see http://orafaq.com/faq/what_is_rac_ops).

> How can the server be setup in a cluster for load-balancing and failover
> like perhaps OPS?

As I understand it, RAC implements a kind of multi-master replication
by sharing database files across multiple machines in conjunction with
a distributed lock manager.

I believe the closest you get to this scheme with PostgreSQL is
PGCluster-II (see
http://www.pgcon.org/2007/schedule/events/6.en.html), which implements
a shared-disk replicaton system. However, I believe the project is
still under development.

There are other ways to implement load-balancing and failover, using
tools such as Slony, pgpool and Skype's londiste. There have been a
dozen threads this year on this topic; I suggest you scan the mailing
list archives.

> How does the Postges solution compare to an Oracle? MSSQL? MySQL solution?

Does MySQL have anything resembling RAC?

Alexander.

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Michael Glaesemann 2007-08-07 17:30:58 Re: finding reusable ids
Previous Message mr19 2007-08-07 16:23:23 UPDATES hang every 5 minutes