Re: Table Partitioning in Postgres:

On Wed, 19 Feb 2003, Christopher Browne wrote:

> In an attempt to throw the authorities off his trail, greg(at)CopelandConsulting(dot)Net (Greg Copeland) transmitted:
> > While I have read that many expect serial ATA to seriously challenge
> > SCSI I honestly have no idea where the rhetoric stops and reality
> > begins. I'd hazard a guess we'll really not know the whole truth until
> > samples become widely available from multiple sources including drives,
> > drivers, and host interfaces.
>
> I'd see Serial ATA being a "serious challenger" if it allowed you to
> cheaply build some "embarrassingly-parallel" RAID servers where you
> were able to get 16 drives hooked up, each on its own ATA 'bus' for
> about the price of 4-way SCSI.
>
> But the intended market is doubtless the super-price-conscious market,
> which naturally leads to the overall quality of the results being
> pretty compromised.

These guys:

http://www.ide-raid.com/

will probably make use of serial ata, since they're already making some
very nice, very fast IDE based RAID devices, and very dense (12 disk 3U
units.)