Alan Eugene Davis wrote: > > This is about a problem that shouldn't have happened, but would have > been easier to recover from if the system were set up a bit > differently. Just US$0.02 worth... > > I upgraded to lilo 17.2, and (first time around) said no to the two > questions from the script (I don't remember clearly what this message > was, something to do with setting up the boot record). Then I could > not reboot. The bootup screen said LI and stopped there. > Yeah, the script is not overly clear on that point. I also had my primary partition on /dev/hdb1 but wanted lilo to install in the boot secotr of /dev/hda, whcih was not possible in the script.
You probably installed the new lilo in the boot sector, but didn't re-run it before re-booting. > I had to use the ?1.1 boot and root disks. I stumbled over a couple > of steps. I didn't realize that I would have to have installed the > kernel before making a new boot disk. I didn't remember the name of > my new kernel, so I couldn't directly use that. Anyway, somehow in > the process, I walked all over the lilo.conf setup. > That probably happened when you installed the new lilo. > ALso, I tried to "execute a shell". For some reason I didn't bother > to investigate, the setup and the shell on the boot root disks could > only see a subset of /etc . > Yes. the shell capabilities from the boot disks are very limited. It would be a good idea to have at least: statically linked versions of cp, mv, ls, mount/umount, as well as ext2fs filesystem support (mine do not have this). The idea is to have a minimal rescue-disk capability in these disks. The slackware distribution had this (does it still?). Obviously there are limitations because of size restrictions. Maybe instead someone could put together a rescue disk set. -- David L. Johnson [EMAIL PROTECTED] [EMAIL PROTECTED] Department of Mathematics http://www.lehigh.edu/~dlj0/dlj0.html Lehigh University 14 E. Packer Avenue (610) 758-3759 Bethlehem, PA 18015-3174
