Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 31 Mar 2003 08:47:41 +0930
From:      Greg 'groggy' Lehey <grog@FreeBSD.org>
To:        Mattias Pantzare <pantzer@ludd.luth.se>
Cc:        Lukas Ertl <l.ertl@univie.ac.at>
Subject:   Re: vinum performance
Message-ID:  <20030330231741.GG1861@wantadilla.lemis.com>
In-Reply-To: <3E87204C.5060304@ludd.luth.se>
References:  <20030330125138.K23911@leelou.in.tern> <3E870CC7.5000204@mac.com> <20030330175605.E23911@leelou.in.tern> <3E87204C.5060304@ludd.luth.se>

next in thread | previous in thread | raw e-mail | index | archive | help

--L+ofChggJdETEG3Y
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

[Format recovered--see http://www.lemis.com/email/email-format.html]

Alternate long/short.

On Sunday, 30 March 2003 at 18:50:20 +0200, Mattias Pantzare wrote:
> Lukas Ertl wrote:
>
>> Ok. But I still don't understand why RAID 5 write performance is _so_ ba=
d.
>> The CPU is not the bottle neck, it's rather bored. And I don't understand
>> why RAID 0 doesn't give a big boost at all. Is the ahc driver known to be
>> slow?
>
> To do a RAID 5 write you do this:
> 1. Read the old data on the blocks that you will write to.
> 2. Read the coresponding parity data.

These two go in parallel.

> 3. Write the new data.
> 4. Write the new parity.

So do these.

> For RAID0 to be faster than a singel disk you have to do reads that
> match or are larger than a stripe. Or have several processes
> generate the IO.

Several processes is the obvious one.  The random seeks test does 3
processes (and you'll note that in this one, RAID-5 exceeds the
performance of a single disk), but that's still not enough.  rawio
uses 8 processes by default, but you can change it.

Greg
--
When replying to this message, please take care not to mutilate the
original text. =20
For more information, see http://www.lemis.com/email.html
See complete headers for address and phone numbers

--L+ofChggJdETEG3Y
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.0 (FreeBSD)

iD8DBQE+h3sVIubykFB6QiMRAquvAJ4uH3qKnMGwYZj17kVI6zkPMbGXqACgg4BQ
75db7faOifzEWykOk3ehIWI=
=az3m
-----END PGP SIGNATURE-----

--L+ofChggJdETEG3Y--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030330231741.GG1861>