MBR in disaster/recovery

[Clark, Patricia]

I had a system with RHEL3 WS U6 with a disk drive that was going bad.
Fortunately, regular backups via Legato had been performed, so full
recovery really was not an issue.  I did, however, want to take the
opportunity to adjust the sizes of a couple of the partitions to better
balance the space usage.  This is what I did:
1.  minimal install of RH OS with "new" partitioning
2.  minimal config to set up for restore from tape
3.  restore from tape - excluding /etc/fstab using overwrite option for
duplicate files.
4.  reboot

Somewhere in making the partitioning mods, I confused grub and/or the
master boot record.  The system was in a permanent reboot loop with
"GRUB" flashing each time it recycled.  I'm thinking that there are
additional files that I should not have recovered and I am unsure what
they are.

5.  upgrade to RHEL4 WS U2 and create a new MBR.  All is well.

As the system was slated for the upgrade anyway, this was not a problem,
just the additional time that my user was without their system wasn't
good.  What should I have done to have the new disk partitioning and
have the system functionally recovered?

Patti Clark
Unix System Administrator
Office of Scientific and Technical Information