Thank you.
kk.cheung
Hmm ... first off -- I would suggest that you compare the EEPROM
contents of the two machines. I suspect in particular that one is set
to boot from disk0, and the other from disk1.
Use "printenv" from the openboot prom level.
In particular, the following setting may be different on the
two machines:
boot-device=disk
That one will boot from disk 0, while a setting of "disk1" will boot
from the other disk slot in the box, and other settings could be
expecting to boot from external disks.
It also could be set to boot from the net, or CD-ROM, or other
possible options.
Compare *all* of the settings, and if you find any different,
figure out whether they make a difference to the boot process.
Good luck,
DoN.
--
Email: <dnic...@d-and-d.com> | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---
>1. I have program running on Sun Ultra 1 with Solaris 8.
>2. I have to move the hard disk from Ultra 1 to Sun Ultra 1 Creator.
>3. When I move the hard disk from Ultra 1 to the Ultra 1 Creator, the
>Creator could not read the disk content to do an Auto Boot. Manual
>boot also could not read the disk content.
>4. I had tried the same hard disk with serveral Sun Ultra 1 and Sun
>Ultra 1 Creator. The result was that all U1 could boot but all U1E
>could not boot.
The system should be able to boot from the disk just fine.
But the booting will fail at some point with "failed to mount root"
or "failed to fsck" or what not.
The reason for this is that the U1 and the U1E use a different SCSI
controller (esp vs fas). So the disk paths used by Solaris are
completely different and even the controller numbers will be different.
If you have a merged / & /usr this is fairly easy to recover from;
if not, than it's much harder.
What may work is replacing /etc/path_to_inst with a file containing
just:
#path_to_inst_bootstrap_1
and touching /reconfigure before shutting the U1 down and moving the
disk to a U1E.
Casper