Re: Unable to kill processes in D-state

From: Jan Hudec (bulb@ucw.cz)
Date: Sun Oct 06 2002 - 12:43:36 EST


On Sun, Oct 06, 2002 at 07:09:32PM +0200, Thomas Lang?s wrote:
> Jan Hudec:
> > If the shares were successfuly reloaded, then the processes should wake
> > up. If they don't, it's a bug in NFS.
>
> They never wake up, and it happens every time.
>
> > Try to reproduce it (ie. reboot some machine, let it start everything
> > and then restart the autofsd and see if processes lock up) and then talk
> > to NFS maintainers about that.
>
> As I said above, it happens every time we encounter this, ie. it's a bug
> that easy to reproduce (since I added nfs@lists.sourceforge.net to the
> CC-list, I'm going to write some of what's already said in this thread).
>
> Problem:
> Processes entering D-state is unkillable. We have a problem with this
> everytime we restart autofs (which automounts quite a few NFS-shares
> on campus), ie. on our samba-boxes smbd hangs forever after this
> (in D-state). Samba still works, it's just that all the D-state processes
> is unkillable and will remain that way untill we reboot the computer.
> Every D-state process increases the load on the machine, and one of our
> 2-CPU intel-boxes currently remains at 430 (which extremly high for such
> a box).
>
> Solution:
> ? :)

You should probably provide detailed info on your setup (computers,
kernel versions, automounter version etc. - as SubmitingBugs in kernel
source suggests). (And maybe change subject to start with NFS PROBLEM,
so more NFS people notice.)

-------------------------------------------------------------------------------
                                                 Jan 'Bulb' Hudec <bulb@ucw.cz>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Oct 07 2002 - 22:00:55 EST