Still having NFS problems

GoogleGroups
1

I got the NFS server installed on my BeagleBone with no problem and the
server came right up. No problem with the shell. I can move all over
the BB from this computer.

I am having a problem with the Nautilus file manager and some other
utilities. I have the BB mounted as /media/bb. Using file manager I
can navigate to /media/bb. I will get one directory listing of the
root. Any further actions locks up the program. Usually the shell
won't work after that either.

Nothing gets written to dmesg on the Ubuntu machine but I see this right
after a boot of the BB.

  2.040222] irda_init()
[ 2.040313] NET: Registered protocol family 23
[ 2.070770] IrCOMM protocol (Dag Brattli)
[ 2.147766] Bluetooth: RFCOMM TTY layer initialized
[ 2.147796] Bluetooth: RFCOMM socket layer initialized
[ 2.147827] Bluetooth: RFCOMM ver 1.11
[ 2.316558] EXT4-fs (mmcblk0p2): re-mounted. Opts: (null)
[ 3.818328] Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
[ 4.072204] svc: failed to register lockdv1 RPC service (errno 97).
[ 4.075592] NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state
recovery directory
[ 4.077056] NFSD: starting 90-second grace period
[ 4.213745] ip_tables: (C) 2000-2006 Netfilter Core Team
[ 4.301513] nf_conntrack version 0.5.0 (3926 buckets, 15704 max)
[ 5.266418]
[ 5.266448] CPSW phy found : id is : 0x7c0f1
[ 5.267211] PHY 0:01 not found
[ 5.307250] ADDRCONF(NETDEV_UP): eth0: link is not ready
[ 6.514862] ADDRCONF(NETDEV_UP): usb0: link is not ready
[ 8.257781] PHY: 0:00 - Link is Up - 100/Full
[ 8.257995] ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[ 18.847076] eth0: no IPv6 routers present

The lines starting at "installing knfsd..." look suspicious but I can't
make any sense of what the messages are saying. I made sure that

/var/lib/nfs/v4recovery

was present (it wasn't so I made it) but other than that I don't know
what to do next.

Help!

Incidentally, how can I get rid of that last line with the 10 second
wait for IPv6 router check?

Thanks,
John

2

________________________________________
From: beagleboard@googlegroups.com [beagleboard@googlegroups.com] on behalf of NeonJohn [jgd@neon-john.com]
Sent: Monday, February 04, 2013 9:50 PM
To: beagleboard@googlegroups.com
Subject: [beagleboard] Still having NFS problems

[snip]

I am having a problem with the Nautilus file manager and some other
utilities. I have the BB mounted as /media/bb. Using file manager I
can navigate to /media/bb. I will get one directory listing of the
root. Any further actions locks up the program. Usually the shell
won't work after that either.

[snip]

Sounds like you are mounting the NFS mount as root from your Linux development machine. Try changing your /etc/exports to something like this:

/exported/dir 192.168.0.0/255.255.255.0(rw,no_root_squash,sync)

3

________________________________________
From: beagleboard@googlegroups.com [beagleboard@googlegroups.com] on behalf of NeonJohn [jgd@neon-john.com]
Sent: Monday, February 04, 2013 9:50 PM
To: beagleboard@googlegroups.com
Subject: [beagleboard] Still having NFS problems

[snip]

I am having a problem with the Nautilus file manager and some other
utilities. I have the BB mounted as /media/bb. Using file manager I
can navigate to /media/bb. I will get one directory listing of the
root. Any further actions locks up the program. Usually the shell
won't work after that either.

[snip]

Sounds like you are mounting the NFS mount as root from your Linux development machine. Try changing your /etc/exports to something like this:

Yes, I am exporting root on my dev system.

/exported/dir 192.168.0.0/255.255.255.0(rw,no_root_squash,sync)

OK, I exported just one volume and the problem remains the same. When I
do a

/etc/init.d/nfs-kernel-server restart

on the dev machine

I get the following in syslog

Feb 4 22:22:01 neonjohn rpc.mountd[32545]: Caught signal 15,
un-registering and exiting.
Feb 4 22:22:01 neonjohn kernel: [168010.372232] nfsd: last server has
exited, flushing export cache
Feb 4 22:22:02 neonjohn kernel: [168011.462230] NFSD: Using
/var/lib/nfs/v4recovery as the NFSv4 state recovery directory
Feb 4 22:22:02 neonjohn kernel: [168011.462250] NFSD: starting
90-second grace period
Feb 4 22:22:02 neonjohn rpc.mountd[307]: Version 1.2.5 starting
Feb 4 22:22:03 neonjohn kernel: [168012.701733] nfs: server
192.168.1.117 OK

when I tried to navigate into /media/bb, syslog logged this message

Feb 4 22:25:08 neonjohn kernel: [168198.112038] nfs: server
192.168.1.117 not responding, still trying

I executed

/etc/init.d/nfsserver restart

on the BB. Dev system syslog registered this:

Feb 4 22:28:12 neonjohn kernel: [168381.759176] nfs: server
192.168.1.117 OK

I remounted /media/bb. Nothing in syslog. Also no response this time,
not even once.

I've tried this from two different computers running different versions
of Ubuntu and got the same results. It seem apparent to me that the BB
NFS server is broken but I don't know what to do from here.

Do you (or anyone else) know where the non-kernel sources for thing such
as nfsd are stored. I'd like to get hold of them and go poking around a
bit.

one last question. With a properly working system, shouldn't I be able
to export root? I do that on my dev, lab and laptop machines with no
problems at all, all running Ubuntu, of course.

Thanks,
John

4

I remounted /media/bb. Nothing in syslog. Also no response this time,
not even once.

I've tried this from two different computers running different versions
of Ubuntu and got the same results. It seem apparent to me that the BB
NFS server is broken but I don't know what to do from here.

Do you (or anyone else) know where the non-kernel sources for thing such
as nfsd are stored. I'd like to get hold of them and go poking around a
bit.

one last question. With a properly working system, shouldn't I be able
to export root? I do that on my dev, lab and laptop machines with no
problems at all, all running Ubuntu, of course.

Thanks,
John

Generally speaking it is not a good idea to export the root of a filesystem of a machine you are developing on, local applications get in the way and sometimes overwrite entries on the remote system. Based on your architecture you might be able to implement an alternative file sharing mechanism such as sshfs, a secure replacement for NFS:

apt-get install sshfs
sshfs root@192.168.51.205:/mnt /mnt/bb

This uses the standard SSH server on the remote machine, no need for any configuration other than that to create a securely tunneled filesystem export over SSH.