Hi everyone,
Here's my problem. I am adding new nodes to an existing cluster using
OSCAR 4.1. The old nodes and the head node are AMD uni-processor boxes
with 10/100 ethernet. They work as is with the standard client image
build. But because the new boxes have Gigbit ethernet, I had to
recompile the kernel and build a new client image for the 2nd set of
nodes. I have used kernel_picker (although I'm not 100% sure of what
this is doing) and have attempted to network boot the new nodes. I am
getting a kernel image not found on the client and the following on the
server:
Jun 6 19:47:24 cerebellum in.tftpd[7093]: RRQ from 192.168.1.113
filename kernel
Jun 6 19:47:24 cerebellum in.tftpd[7094]: RRQ from 192.168.1.113
filename kernel.cbt
Jun 6 19:47:24 cerebellum in.tftpd[7095]: RRQ from 192.168.1.113
filename kernel.0
I've been dancing around this for a few days and can't seem to find the
problem. If the answer is obvious, please enlighten me! If not, I have
a few questions.
During the PXE boot process, is it the initrd.img that actually gets
booted on the client and then that starts the installation process?
How does the client image builder (in the Oscar Install Wizard) know
which kernel to use? ie. there can be multiple kernel images in /boot.
It seems that the installer builds the clients to boot using lilo. Is
it possible to tell it to use grub instead?
Well that's it for now. I'll probably be bothering everyone again
later, but thanks in advance for your help.
John Hayden
-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games. How far can you shotput
a projector? How fast can you ride your desk chair down the office luge track?
If you want to score the big prize, get to know the little guy.
Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Oscar-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/oscar-users
