Re: Large Databases redux

On Wed, Mar 21, 2012 at 02:58:43PM -0700, John R Pierce wrote:
> On 03/21/12 2:18 PM, Jason Herr wrote:
> >I have my own theories based on what I've read and my puttering.
> >I think I can get away with a disk for the OS, disk for the WAL,
> >disk for the large table (tablespaces) and a disk for the rest.
> >And when I say disk I mean storage device. I'm thinking RAID1 15k
> >disks for each set but the databases and then raid 10 or VERY
> >large disks.
>
> I think you're better off just making one big raid10 out of all the
> disks and putting everything on it, maybe in different file systems
> to seperate out file fragmentation. this way the IO workload is
> evenly distributed across all the disks.

That, and a good RAID controller with BBU cache will go a long way to
relieving the pain of fsync.

Have a nice day,
--
Martijn van Oosterhout <kleptog(at)svana(dot)org> http://svana.org/kleptog/
> He who writes carelessly confesses thereby at the very outset that he does
> not attach much importance to his own thoughts.
-- Arthur Schopenhauer