Re: NAS, SAN or any alternate solution ? - Mailing list pgsql-performance

From Rod Taylor
Subject Re: NAS, SAN or any alternate solution ?
Date
Msg-id 1090344321.7056.105.camel@jester
Whole thread Raw
In response to NAS, SAN or any alternate solution ?  (bsimon@loxane.com)
Responses Re: NAS, SAN or any alternate solution ?  (Andrew Hammond <ahammond@ca.afilias.info>)
List pgsql-performance
> Would NAS or SAN be good solutions ? (I've read that NAS uses NFS
> which could slow down the transfer rate ??)
> Has anyone ever tried one of these with postgresql ?

I've used both a NetApp and Hitachi based SANs with PostgreSQL. Both
work as well as expected, but do require some tweeking as they normally
are not optimized for the datablock size that PostgreSQL likes to deal
with (8k by default) -- this can make as much as a 50% difference in
performance levels.

For a NAS setup, be VERY careful that the NFS implementation you're
using has the semantics that the database requires (do plenty of failure
testing -- pull plugs and things at random). iSCSI looks more promising,
but I've not tested how gracefully it fails.

Have your supplier run a bunch of benchmarks for random IO with 8k
blocks.

One side note, SANs seem to be very good at scaling across multiple jobs
from multiple sources, but beware your Fibre Channel drivers -- mine
seems to spend quite a bit of time managing interrupts and I've not
found a way to put it into a polling mode (I'm not a Linux person and
that trick usually happens for me on the BSDs).



pgsql-performance by date:

Previous
From: "Steinar H. Gunderson"
Date:
Subject: Re: Odd sorting behaviour
Next
From: andrew@pillette.com
Date:
Subject: Unbearably slow cascading deletes