Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 24 Jun 2004 16:00:09 +0200
From:      =?ISO-8859-1?Q?S=F8ren_Schmidt?= <sos@DeepCore.dk>
To:        Daniel Eriksson <daniel_k_eriksson@telia.com>
Cc:        'FreeBSD Current' <freebsd-current@freebsd.org>
Subject:   Re: Multiple choices for RAID-0, best performance?
Message-ID:  <40DADE69.30702@DeepCore.dk>
In-Reply-To: <!~!UENERkVCMDkAAQACAAAAAAAAAAAAAAAAABgAAAAAAAAA0VcX9IoJqUaXPS8MjT1PdsKAAAAQAAAArboJgGwHU0apkIeCcojdKwEAAAAA@telia.com>
References:  <!~!UENERkVCMDkAAQACAAAAAAAAAAAAAAAAABgAAAAAAAAA0VcX9IoJqUaXPS8MjT1PdsKAAAAQAAAArboJgGwHU0apkIeCcojdKwEAAAAA@telia.com>

Next in thread | Previous in thread | Raw E-Mail | Index | Archive | Help
Daniel Eriksson wrote:

> Also, will the RAID-0 performance (transfer speed and/or resource usage) of
> ataraid be affected a lot depending on what controller is used? For example
> would there be any performance difference between ataraid on a HPT374-based
> card (such as the Highpoint RocketRAID 454) and one of the new Promise cards
> (which we have really good drivers for afaik).

Well, I havn't measured the performance diff between HPT and Promise 
lately, but I'm pretty sure Promise wins due to much lower command overhead.

However in the not to far future this wil change *significantly*.
ataraid is in the learning phase of using the command sequencer in the 
Promise (and Marvell) chips which will boost performance by having the 
controllers do some of the RAID work (in case of the pdc20621 chips in 
can even do HW XOR for RAID5).

Needless to say that it will only work with ataraid, as this needs 
intimate knowledge of the ATA controller workings to be used effectively.


-- 
-Søren



Want to link to this message? Use this URL: <http://docs.FreeBSD.org/cgi/mid.cgi?40DADE69.30702>