Re: Hardware Recommendations

From: "Mitch Vincent" <mvincent(at)cablespeed(dot)com>
To: "PostgreSQL-General" <pgsql-general(at)postgresql(dot)org>
Subject: Re: Hardware Recommendations
Date: 2001-07-25 22:13:46
Message-ID: 020101c11557$13192e80$1251000a@Mitch
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

The dual AMD board(s? I only know of one..) are *really* new, so I'd be
careful.... They sure do look nice though and aren't that expensive compares
to those quad Xeons...

A Note on Disks :
I respectfully disagree with Steve's "disk speed isn't terribly
important" -- it can be (and is totally dependent on what you're doing) and
since you have the on-board U160 with all the dual Athlon boards I've seen,
you might as well take full advantage of it.. That and you know Seagate has
those new 15,000 RPM U160 drives out *evil grin*..

A note on RAM :
Note that those dual Athlon boards take registered DDR RAM, which is a
little more expensive than your run-of-the-mill SDRAM (about twice(ish) as
much according to some numbers I just looked up). I wish I had something
going that needed that much horsepower so I could justify buying one!

Good luck!

> If I were to set up another machine now, I'd use a dual Athlon. For
> not much more than the dual P3 would cost you, you could have a 2.4 GHz
> machine with DDR RAM - and here's the real kicker, the dual Athlon boards
> have a seperate 266 MHz bus to *each* processor, and a 266 MHz bus to the
> RAM - so it's like each CPU having a 256 MHz bus to the chipset, and a 133
> MHz bus to the RAM - about 5 times the bandwidth that a Xeon has in a
> quad-configuration, and twice what a P3 would have. It's a win in every
> regard over the Xeons and P3's, in my opinion.
>
> As for the disk, the disk speed isn't terribly important, provided that
> you're not using fsync() and have plenty of RAM for caching. On our quad
> Xeon with 1.5 GB's, the disk lights only blink very occasionally.
> However, we do use RAID 5 + hot spare for redundancy, so that a disk
> failure won't take us down.

-Mitch

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Stephen Robert Norris 2001-07-25 22:19:13 Re: postgres deadlocks
Previous Message Steve Wolfe 2001-07-25 21:53:07 Re: Hardware Recommendations