AFAIK google uses thousands of pcs with IDE drives. So going the multiple
PC way can work.
Not sure how that would be implemented for postgresql. It seems simple to
support _many_ read only queries at a time using many pcs. But how would
one speed up a few large parallel queries that way?
Cheerio,
Link.
At 08:08 AM 4/23/02 +0900, Curt Sampson wrote:
>On Mon, 22 Apr 2002, Francisco Reyes wrote:
>
> > IDE's may have good thoughput, but their seek times can't compete with top
> > of the line SCSI.
>
>Dollar for dollar, IDE has far better performance. Even including the
>cost of extra controllers (because you need one controller for each
>drive), I can get four high-end IDE drives for the price of one SCSI
>drive. There's no way any SCSI drive is going to do as many I/Os per
>second as four good IDE drives.
>
>As well, seek time isn't always important. For your log disk, for
>example, you care more about high sequential write speed.
>
> > Exactly.. since it won't be easy for you to find the best distributions
> > then it may be worth getting better hardware. :-)
>
>What makes you think it won't be easy, in my case?
>
>cjs