NFS booting suddenly stops working (mount.nfs an incorrect mount option.....)

[panachoi]

I've been booting via NFS without troubles for some time. Now however, I'm unable to NFS boot *ANY* kernel, even previously working ones (ie the ones from sheeva.with-linux.com). I'm running Debian Sid/Squeeze. Now, no matter which kernel I'm NFS booting, I'm getting the following:

Code:

[    8.214746] scsi1 : usb-storage 1-1.1:1.0
[    8.220043] usbcore: registered new interface driver usb-storage
[    8.226117] USB Mass Storage support registered.
done.
Activating swap...done.
mount.nfs: an incorrect mount option was specified
Checking root file system...fsck from util-linux-ng 2.17.2
done.
mount.nfs: an incorrect mount option was specified
Loading kernel modules...done.
/etc/init.d/mtab.sh: 178: cannot create /etc/mtab: Read-only file system
Cleaning up ifupdown....
Activating lvm and md swap...done.
Checking file systems...fsck from util-linux-ng 2.17.2
done.
Setting up networking....
Mounting local filesystems...can't create lock file /etc/mtab~343: Read-only file system (use -n flag to override)
failed.
Activating swapfile swap...done.
Cleaning up temporary files.../lib/init/bootclean.sh: 186: cannot create /tmp/.clean: Read-only file system
[    9.247259] scsi 1:0:0:0: Direct-Access     Generic  STORAGE DEVICE   9909 PQ: 0 ANSI: 0
df: Warning: cannot read table of mounted file systems: No such [    9.262021] scsi 1:0:0:1: Direct-Access     Generic  STORAGE DEVICE   9909 PQ: 0 ANSI: 0
file or directory
Setting kernel variables ...done.
Configuring network interfaces...[    9.382591] sd 1:0:0:0: [sda] Attached SCSI removable disk
[    9.389365] sd 1:0:0:1: [sdb] 3964928 512-byte logical blocks: (2.03 GB/1.89 GiB)
[    9.398855] sd 1:0:0:1: [sdb] Write Protect is off
[    9.403734] sd 1:0:0:1: [sdb] Assuming drive cache: write through                       
[    9.416813] sd 1:0:0:1: [sdb] Assuming drive cache: write through                       
[    9.422954]  sdb: sdb1                                                                 
[    9.435062] sd 1:0:0:1: [sdb] Assuming drive cache: write through                       
[    9.441271] sd 1:0:0:1: [sdb] Attached SCSI removable disk                             
done.
Starting portmap daemon....
Starting NFS common utilities: statd failed!
Cleaning up temporary files.../lib/init/bootclean.sh: 186: cannot create /tmp/.clean: Read-only file system
/etc/init.d/bootmisc.sh: 63: cannot create /var/run/utmp: Read-only file system
startpar: service(s) returned failure: mtab.sh mountall-bootclean.sh nfs-common mountnfs-bootclean.sh bootmisc.sh ... failed!

I have no problem mounting the root filesystem R/W when booted from NAND+SD, so its not a server problem. I've even went as far as making an entirely new NFS root, with the same result.

I'm wondering if anyone has seen this, or if somehow there's somekind of bug in some underlying package in Sid/Squeeze that affects NFS root....

[panachoi]

Just in case anyone runs into this oddity, I've "solved" the problem. I rebuilt my NFS root filesystem with debootstrap, and everything worked again. At least until I altered the client /etc/fstab file with the following entry:

amanita:/usr/virtimages/plugroot       /       nfs     rw,nolock 0 0

Bam, same error again. When I comment out this line, everything works as expected. This used to work just fine, so I'm sort of wondering what changed in the handling of mounting filesystems at boot that  made this stop working. After all, this is same as given on the command line boot and delivered via DHCP. It just seems that the boot process is confused by this. Its a bit ugly without it, but luckily there are only a couple of fileservers here at home whcih could possibly hold the root filesystem for this "box"..