[Top] [All Lists]

Re: Stale XFS mount for Kernel

To: "Ngo, Andrew" <andrew.ngo@xxxxxxxx>
Subject: Re: Stale XFS mount for Kernel
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Mon, 13 Oct 2008 14:59:39 +1100
Cc: xfs@xxxxxxxxxxx, "Johnson, Je" <je.johnson@xxxxxxxx>
In-reply-to: <8604545CB7815D419F5FF108D3E434BA3BD626@xxxxxxxxxxxxxxxxxxxxx>
Mail-followup-to: "Ngo, Andrew" <andrew.ngo@xxxxxxxx>, xfs@xxxxxxxxxxx, "Johnson, Je" <je.johnson@xxxxxxxx>
References: <8604545CB7815D419F5FF108D3E434BA3BD626@xxxxxxxxxxxxxxxxxxxxx>
User-agent: Mutt/1.5.18 (2008-05-17)
On Sun, Oct 12, 2008 at 10:10:54PM -0400, Ngo, Andrew wrote:
> Hi,  
> This is my first attempt to report a stale mount problem.  If this is
> not the relevant mail list, please let me know.  I suspect that my
> problem is xfs related, thus I select this mailing list.
> I have a system that recently upgraded it's kernel to  The
> system's previous kernel was 2.6.18.-8.el5.  With the previous kernel, I
> have no problem with executing the following commands when the system is
> in operation.  Note that the /mnt/mountpoint is mounted on a XFS file
> system.
> mount -o rw,remount /mnt/mountpoint
> mount -o ro,remount /mnt/mountpoint
> With kernel, however, the mount command may be hang when
> changing from rw to ro.  When this happen, system is very sluggish.

That's because the remount,ro is actually flushing all the dirty
data to disk in 2.6.25. There is also a check to catch when the
remount,ro fails to clean everything that is dirty - that's probably
what is triggering. Do you have an oops in the dmesg log when
mount hangs?


Dave Chinner

<Prev in Thread] Current Thread [Next in Thread>