Samuel Flory wrote: > Carl-Daniel Hailfinger wrote: > >> >> Funny - I had a promise ataraid installation that no version of lilo >> could boot. Tried grub once and it worked. Since then I avoid lilo. >> > That seems to be the reverse of my experince. Luckily, both of us have a bootloader that "works for me (TM)". >> However, we are now talking about md devices. Unfortunately I have no >> idea how grub behaves there but I heard that a new grub is able to >> handle that without problems. >> > Maybe there is, but it's not in RH 9. (Or the installer doesn't know > how to configure it.) In which the easier path for people is to check > the use lilo box durning install. > >> Testing grub is always easy. Simply use a grub boot floppy from >> somewhere and try to boot your kernel from there. If it works, it also >> should when installing grub to harddisk. > > You're implying that it's easy to install grub to both drives. If > there is an easy way I don't know about it. Neither does radhat: > http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=55484 AFAICS, the problem exists only with grub-install. If you install grub manually (using grub itself and not the grub-install script), it seems to work. Or did I miss something written down in the bug? Carl-Daniel
