linuxgeek: Re: [nslu2-linux] NFSD: unable to find recovery directory /var/lib/nfs/v4recovery

slackline wrote:
> Hi,
>
> Unfortunately my slug got knocked off the shelf when my (ill) cat failed to jump onto the shelf. As a result it had a power-cycle without correct shutdown and I now find I'm having trouble exporting (and mounting from a client) my NFS directories.
>
> I see in dmesg....
>
> eth0: link up, 100Mbps, full-duplex, lpa 0x45E1
> RPC: Registered udp transport module.
> RPC: Registered tcp transport module.
> Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
> NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
> NFSD: unable to find recovery directory /var/lib/nfs/v4recovery
> NFSD: starting 90-second grace period
>
Red herring. SlugOS 5.3 doesn't do NFS v4 (it supports V3 and V2), so
this message is just noise (we probably should have disabled V4 support
altogether, but the effort to do so didn't seem worth the time, since
it's just a cosmetic message).

> and from the client I get permission denied on all directories...
>
> # /etc/init.d/netmount restart
> * Caching service dependencies ... [ ok ]
> * Unmounting network filesystems ... [ ok ]
> * Mounting network filesystems ...
> mount.nfs: access denied by server while mounting slug:/mnt/torrents
> mount.nfs: access denied by server while mounting slug:/mnt/albums
> mount.nfs: access denied by server while mounting slug:/mnt/music
> mount.nfs: access denied by server while mounting slug:/mnt/video
> mount.nfs: access denied by server while mounting slug:/mnt/pics
> mount.nfs: access denied by server while mounting slug:/mnt/doc
> mount.nfs: access denied by server while mounting slug:/mnt/work1
> mount.nfs: access denied by server while mounting slug:/mnt/work2
> mount.nfs: access denied by server while mounting slug:/mnt/ref
> * Could not mount all network filesystems [ !! ]
>
>
> I found a suggestion of adding the 'nolock' option to /etc/exports but I'm told this isn't recognised (and its not listed in the man page for /etc/exports either).
>
> Currently my /etc/exports looks like...
>
> # /etc/exports: NFS file systems being exported. See exports(5).
> ## /usr/portage
> /mnt/portage 192.168.1.0/255.255.255.0(rw,async,subtree_check,no_root_squash)
> ## 250Gb LaCie
> ## pics1
> /mnt/torrents 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## music1
> /mnt/albums 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## 500Gb LaCie
> ## music2
> /mnt/music 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## video
> /mnt/video 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## pics2
> /mnt/pics 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## doc
> /mnt/doc 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## misc
> /mnt/work1 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## work
> /mnt/work2 192.168.1.0/255.255.255.0(rw,async,subtree_check)
> ## ref
> /mnt/ref 192.168.1.0/255.255.255.0(rw,async,subtree_check)
>
>
> The versions of NFS (and modules) I have installed are...
>
> # opkg list_installed | grep nfs
> kernel-module-nfs - 2.6.27.8+svnr1085-r3 -
> kernel-module-nfsd - 2.6.27.8+svnr1085-r3 -
> nfs-utils - 1.1.2-3 -
>
>
> I get the impression that there is a lock on NFS exports due to the unclean power-cycle, but have no idea how to get round this.
>
There's no lock involved; there is no state maintained for NFS mounts
themselves (and there's actually no locking maintained unless you are
running the lockd, but even so it won't prevent a mount from happening).

Look for a permissions issue. I would wonder if, at the time your
exports are being processed, those mount points in /mnt are actually
mounted yet?

-Mike (mwester)