Home

Problems installing on NFS server

joeuser

Usuario
Amigo OpenSPA
7 Feb 2015
21
9
0
I am usually testing new images from my NFS server and tried to do the same with this new image. I extracted all the files from the e2jffs2.img and put the uImage file in the boot directory. After that I was able to use uboot to load the uImage into memory and start an nfs boot from the server. It started booting ok, but at some point it stalled and I received nfs mount errors. After looking at the boot sequence, I found the following:

init runs /etc/init.d/rcS
at the end of rcS, it calls:
exec /etc/init.d/rc S

but inittab sets default runlevel to 3
so init runs the startup scripts for runlevel S which includes
/etc/init.d/ddbootup

and at the end of that script, it removes the ip address set by the kernel, causing the nfs mount to fail...
.
.
.
# Flush eth0 device which is set by kernel cmdline
ip addr flush dev eth0

By commenting out the last line, the system boots fine from an nfs server.

Not sure why rcS is setting runlevel to S?? It looks like enigma2 is launched by init due to the following line in inittab:
gui:3:respawn:/usr/bin/enigma2.sh

So, I am not sure what the intended runlevel of the system is.
 

joeuser

Usuario
Amigo OpenSPA
7 Feb 2015
21
9
0
I see now that I posted in the wrong place. Maybe a moderator can move this to the correct place???