Thursday, 22 August 2013

PXE boot with NFS root is trying to use 'user_xattr' option to mount and failing

PXE boot with NFS root is trying to use 'user_xattr' option to mount and
failing

I'm building a small diskless cluster using PXE boot and a copy of SLES 11
SP1 from a working hard disk install. The kernel fetches over tftp and
boots fine, but during boot I see this:
Mounting root 192.168.1.3:/exports/SLES11sp1
mount -o rw,acl,user_xattr -t nfs 192.168.1.3:/exports/SLES11sp1 /root
mount: wrong fs type, bad option, bad superblock on
192.168.1.3:/exports/SLES11sp1
But nfs does not support 'user_xattr'. If I try the mount myself, without
the user_xattr option, it works fine. Here's my pxelinux.cfg/default file:
# SLES11sp1 boot
label linux
kernel vmlinuz-2.6.32.12-0.7-try2
append root=/dev/nfs nfsroot=192.168.1.3:/exports/SLES11sp1
initrd=initrd-2.6.32.12-0.7-try2 raid=noautodetect
ipappend 1
And here's /etc/fstab on the boot image:
192.168.1.3:/exports/SLES11sp1 / nfs defaults 0 0
proc /proc proc defaults
0 0
sysfs /sys sysfs noauto
0 0
devpts /dev/pts devpts mode=0620,gid=5
0 0
tmpfs /dev/shm tmpfs defaults,size=4G
0 0
I'm pretty sure that if I can just get the kernel to stop trying to use
'user_xattr' when mounting the root filesystem, it'll work fine? I've
already tried
append root=/dev/nfs nfsroot=192.168.1.3:/exports/SLES11sp1,nouser_xattr
initrd=initrd-2.6.32.12-0.7-try2 raid=noautodetect
with no luck.

No comments:

Post a Comment