Ive been doing some testing and come up with this...
I noticed the IO errors starting back up again and the only file that was
being downloaded from the drives was a 50mb data file. I rebooted since
this is the only way I can find to stop the errors from happening. I then
downloaded the same data file again, but no errors were reported.
I don't see how this can be anything other than a bug with either XFS or
LVM, as the same drive access after a reboot causes no problems.
Im still stuck with no ideas on how to fix this problem, Im beginning to
wish I hadn't used XFS :(
Cheers,
Tom
-----Original Message-----
From: Chris Wedgwood [mailto:cw@xxxxxxxx]
Sent: 29 June 2004 01:19
To: Joshua Schmidlkofer
Cc: Tom Shaw; linux-xfs@xxxxxxxxxxx
Subject: Re: LVM and XFS problems - xfs_force_shutdown
On Mon, Jun 28, 2004 at 03:29:31PM -0700, Joshua Schmidlkofer wrote:
> Is there any chance of fixing this so that all the numbers see are
> consistent?
Probably not, the fact it was hex and misread was my fault. I was
tired and not thinking straight, usually it's not a big problems.
> Is it insane to ask? The same problems exists with iptables
> vs. iproute2. iptables takes it in decimal, and iproute2 takes it
> in hex. Those are config tools, and I can take that up with the
> authors, but what about this?
If you feel strongly enough about it post a patch to lkml for
discussion.
--cw
|