What's the expected transaction split (read/write)?
If mostly READs (e.g. SELECTs) then its very, very hard to do better from a performance perspective than Raid 1 with the transaction log on a separate array (physically separate spindles)
I run a VERY busy web forum on a Quadcore Intel box with this setup and it is very fast. Really quite amazing when you get right down to it.
The latest release of the PostgreSQL code markedly improved query optimization, by the way. The performance improvement when I migrated over was quite stunning.
Karl Denninger (karl@denninger.net)
http://www.denninger.net
Robert Powell wrote:
To whom it may concern,
I’m looking for a file server that will give me a high level of redundancy and high performance for a postgres database. The server will be running only postgres as a backend service, connected to a front end server with the application on it.
I was thinking along the lines of a RAID 10 setup with a very fast processor.
Any suggestions would be greatly appreciated.
Bob Powell
The Hotchkiss School
