On Dec 26, 2007 7:36 PM, Chris Hoover <revoohc(at)gmail(dot)com> wrote:
> Is anyone running their production PostgreSQL server on the RHEL Cluster
> software? If so, how is it working for you? My linux admin is looking at
> trying to increase the robustness of our environment and wanting to try and
> eliminate as many single points of failure as possible.
> So, I am looking for any hands on experience in running PostgreSQL in some
> sort of HA cluster.
I was looking for the same information previously this year (
http://archives.postgresql.org/pgsql-general/2007-08/msg00910.php). At that
time I concluded that using drbd for data replication was probably better
choice than using RHCS and shared storage, but afterwards the requirements
for the cluster changed so that drbd was no longer a suitable choice, and we
ended up with building a RHCS cluster with shared disk and gfs.
The cluster has been working as expected and not counting a few minor
annoyances I haven't had any problems with it (this is with RHEL 5.1). My
initial fear was that somehow two postmasters could be accessing the same
set of file but after testing I have not been able to produce this (ie. the
fencing is working). So overall I see no reason not to go with RHCS if it
satisifies your business requirements.
In response to
pgsql-performance by date
|Next:||From: Kevin Grittner||Date: 2007-12-28 15:05:36|
|Subject: Re: Evaluation of PG performance vs MSDE/MSSQL 2000(not 2005)|
|Previous:||From: Tom Lane||Date: 2007-12-27 18:54:28|
|Subject: Re: More shared buffers causes lower performances |