Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 24 Nov 2004 09:02:11 +1030
From:      Greg 'groggy' Lehey <grog@FreeBSD.org>
To:        Ivan Voras <ivoras@fer.hr>
Cc:        current@freebsd.org
Subject:   Re: gvinum panic?
Message-ID:  <20041123223210.GL21905@wantadilla.lemis.com>
In-Reply-To: <41A33907.5020703@fer.hr>
References:  <41A2488F.8030802@fer.hr> <20041122215228.J568@korben.in.tern> <41A33907.5020703@fer.hr>

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

--Mit9XoPEfICDqq/V
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

On Tuesday, 23 November 2004 at 14:20:07 +0100, Ivan Voras wrote:
> Lukas Ertl wrote:
>> On Mon, 22 Nov 2004, Ivan Voras wrote:
>>
>>> I'm consistently getting a system panic, trap #12 in process g_event
>>> when I try:
>>
>> Are you sure that kernel (modules) and userland are in sync?
>
> Thanks, it looks like they've been unsynced for a while :(
> It's ok now.
>
> But, the performance is pretty bad (same with regular vinum). I tried
> setting up graid3 class with same parameters, and got these results:
>
> linear read (1024 MB file, blocks of 32768) .................  38.5 MB/s
> linear write (1024 MB file, blocks of 32768) ................  16.8 MB/s
> random read (1024 MB file, blocks of 32768) .................  41.9 MB/s
> random write (1024 MB file, blocks of 32768) ................   3.4 MB/s
>
>
> with raid5, gvinum or vinum, I get these:
>
> linear read (1024 MB file, blocks of 32768) .................  27.8 MB/s
> linear write (1024 MB file, blocks of 32768) ................   4.2 MB/s
> random read (1024 MB file, blocks of 32768) .................  44.2 MB/s
> random write (1024 MB file, blocks of 32768) ................   1.9 MB/s
>
> (random reads are really random, no clustering or anything).
> My setup is 2 IDE channels with three discs, so one is a slave (and
> contains the parity disc in case of raid3). Do these numbers make
> sense?

Not without clarifying what you're doing.  If you're only doing one
request at a time, there's no way that RAID can help you.  I find it
highly unlikely that all three RAID alternatives generate the same
numbers.

Greg
--
See complete headers for address and phone numbers.

--Mit9XoPEfICDqq/V
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (FreeBSD)

iD8DBQFBo7pqIubykFB6QiMRAgpJAKCfsG6SCRzYCyfT3pTNG56rC0hlzwCcD0Jj
0YLJz+Kw4ZiJA/gxvq4SjIE=
=+Q9O
-----END PGP SIGNATURE-----

--Mit9XoPEfICDqq/V--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20041123223210.GL21905>