[Mageia-dev] Lazy NFS umount problem: kernel or userspace?

[Pascal Terjan]

On Mon, Nov 19, 2012 at 9:39 AM, Colin Guthrie <mageia at colin.guthr.ie>wrote:

> Hi,
>
> For many, many years, I've always done a umount -l  on stale NFS mounts
> (e.g. when resuming in a different location to suspending).
>
> This has always worked, but very recently (i.e. in the last couple
> weeks) this has stopped working.
>
> Anyone know what could be up?
>
> Blino's change to nfs-utils doesn't look like a problem.
>
> The call that's stalling things is stat(), but this never used to be a
> problem with lazy umounts in the past (not sure if it's just stat()
> that's now blocking at a kernel level or of some userspace code now
> calls it when in the past it did not?
>

If your nfs mount is dead I would expect stat to hang, so I guess it wasn't
called in the past.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: </pipermail/mageia-dev/attachments/20121119/bb9785c8/attachment.html>