[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Testbed-admins] Can't boot after imageunzipping a customized image



We've created and customized the images described in the
documentation, then imagezipped them back to create the ndz files on
boss.  but whenever we unzip them to another pc to try to boot them
into the new OS, the interactive mode prompt either doesn't
recognize the OS on the drive, or doesn't see it as bootable.  is
the master boot record getting copied correctly?  what information
can I send for you to look at to see?

Also when we unzipped the fedora core 8 image onto the FBSD62+FC8
machine then remade the new image, imagezip couldn't determine the
superblock saw raw compression was needed.  same was true for CentOS
and WinXP, all are being forced to use raw compression.  this is
extremely slow, is there a better way to determine what compression to
use, is there a command line option for imagezip to specify it?

I think all of these problems are related to one thing; the MBR on the
disk is getting whacked somehow. Its my best theory at the moment,
without being able to look at it myself.

So lets do this. Lay down the generic image. Make sure it boots. Then
boot into the admin MFS, and imagezip it with this:

boss> wap node_admin on pcXXX
boss> cd /usr/testbed/images
boss> sudo ssh pcXXX imagezip /dev/$DSK - > foo.ndz

Then lay that image back down:

boss> sudo ssh pcXXX imageunzip - /dev/$DSK < foo.ndz
boss> wap node_admin off pcXXX

If the node does not boot, please upload foo.ndz to your account in
Utah. You will have to put in /proj/AirNet/images cause of quota.

Lbs