Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 18 Oct 2012 11:35:18 +0200
From:      Kristof Provost <kristof@sigsegv.be>
To:        Alexey Tyurikov <alexey.tyurikov@gmail.com>
Cc:        freebsd-fs@freebsd.org
Subject:   Re: ZFS over NFS issue with linux clients
Message-ID:  <20121018093517.GS9028@thebe.jupiter.sigsegv.be>
In-Reply-To: <CANJVYoLWWPDJ0r=RnWiSnUTKnY3AD5K-LL6QkWQecmuqjVLw0Q@mail.gmail.com>
References:  <CANJVYoLWWPDJ0r=RnWiSnUTKnY3AD5K-LL6QkWQecmuqjVLw0Q@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On 2012-10-18 10:50:25 (+0200), Alexey Tyurikov <alexey.tyurikov@gmail.com> wrote:
> I'm looking for people using FreeBSD 9.0 amd64 ZFS over NFS (v3 and v4)
> with Linux clients. My clients (last CentOS6 and Debian6) don't use cache
> for files lying on ZFS over NFS. Tested:
> 
>  $ grep something /nfs_zfs/file-100MB # network activity, client read the
> file over NFS (slow)
>  $ grep something /nfs_zfs/file-100MB # network activity, client read the
> file over NFS once again (slow)
> 
I'm unable to reproduce this behaviour with a Debian 6.0 client (ARM,
because my openrd is the only client I can get to at the moment). That's
a 2.6.32-5-kirkwood kernel.

The server is FreeBSD 9.1-PRERELEASE #91 r241441 (amd64). I can't
easily compare to UFS, because the server is ZFS only.

I see network activity on the first read, but only very little (a single
getattr + reply in fact) on the second read.

Regards,
Kristof




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20121018093517.GS9028>