Re: kernel-nfs messing up a few things

Philippe Troin (phil@fifi.org)
11 Mar 1999 07:43:20 -0800


Harald Kirsch <kir@iitb.fhg.de> writes:

> We have a setup with diskless clients running kernel-nfs at kernel
> 2.2.3. In particular
>
> /
> /usr
> /var
> /tmp
> /etc
>
> and so on are mounted via nfs, some read-only some, like /, /tmp and
> /etc are mounted read-write. The problem we have manifests with the
> following messages on the client:
>
>
> Mar 11 13:12:21 neptun kernel: __nfs_fhget: inode 49316 busy, i_count=2, i_nlink=1
> Mar 11 13:12:21 neptun kernel: nfs_free_dentries: found bin/rxvt, d_count=5, hashed=0
> Mar 11 13:12:21 neptun kernel: __nfs_fhget: inode 49316 still busy, i_count=2
> Mar 11 13:12:21 neptun kernel: __nfs_fhget: killing bin/rxvt filehandle
> Mar 11 13:12:21 neptun kernel: __nfs_fhget: inode 917613 busy, i_count=2, i_nlink=1
> Mar 11 13:12:21 neptun kernel: nfs_free_dentries: found .X11-unix/X0, d_count=13, hashed=0
> Mar 11 13:12:21 neptun kernel: __nfs_fhget: inode 917613 still busy, i_count=2
> Mar 11 13:12:21 neptun kernel: __nfs_fhget: killing .X11-unix/X0 filehandle
>
>
> In particular `killing .X11-unix/X0 filehandle' is uncool, because
> afterwards it is no longer possible to start X clients. No connection is
> possible to the X-server anymore.

8< snip >8

Same problem here.
Diskless client is just unusable in the current 2.2.x state.

Phil.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/