Well you can't use another server emergency boot diskette becos the kernel is different from the server you intend to rescue. However, since you can enter with other users you can then try using the authck command to repair the file corrupted. After you have done that use the fixmog and the integrity command and reboot your system. Try this and let see the outcome.
Thanks for the answer. Finally I travel to server's location (500 miles away !) and see what was happening. Root account was disabled and tty01 too. (override terminal). Because of that, root can't log in, even in maintenance mode.
I started the system from installation CDROM and when prompted for Licences, I press F8 to exit to a shell. Then I create /dev/hd0root (mknod /dev/hd0root b 1 42) and mounted the server's root filesystem.
With this filesystem mounted, I used chroot to use vi on some files. I don't know why these accounts and terminal was disabled, but I remove root password from /tcb/files/auth/r/root and /etc/shadow, delete some code that seems to be locks, unmounted filesystem and reboot. I changed OVEERRIDE in /etc/default/login to /dev/tty03.
Note: when using chroot , there are some issues with commands and files that are expected to be on /etc, for instance. A lot are links to something like /var/opt/K/SCO/Unix/5.0.5Eb/etc or something like this. I you want to edit /etc/shadow of a root filesystem mounted as said before, command line should look like :