It has been my experience that doing an "upgrade" is not a great idea. A
much better idea is to save what you need, wipe out the root and stand
partitions and reinstall from scratch. We do a lot of SCO upgrades, and
have never actually done the in-place upgrade on a real machine. (We've
played with it on our test servers, but decided it was better to
reinstall...)
Good luck,
Fabio
I agree.
Get and use my savefiles script (ftp.jpr.com).
--
JP
I don't care for upgrades. See http://pcunix.com/Unixart/upgrades.html
for a pretty complete treatment, samples of useful scripts, etc.
--
Tony Lawrence
SCO/Linux Support Tips, How-To's, Tests and more: http://pcunix.com
I would rate my try at in-place upgrade 5.0.4 -> 5.0.5 one of my all-time
biggest career panic attacks! A 100-user health-care system.
I'll never forget how carefully I installed it, by CD, following
instructions, ever so closely. I remember taking a deep breath of relief
when the SCO script said something like:
"Installation is complete, Upgrade successful, you must now restart your
system for the upgrade to take effect"
Oh, what a sigh of relief. Then I reboot... ... ... ... ... ... ... it never
reboot again.
Two days later, high blood pressure, and anxious management looking on, I
blew it all away, fresh installed 5.0.5, and restored apps from backup.
I would not touch an in-place upgrade ever again!
I did an upgrade from 5.0.4 to 5.0.5 on a live server. I was young, I
was idealistic, I was full of big ideas to make the world a better
place. I even got in touch with SCO who told me "There will be
absolutely no problems whatsoever".
The moment it went wrong and the subsequent steps I had to take to
sort it out is something that will stay with me forever.
Cancel the weekend (pretend your Aunt died, anything). Then come up
with a plan that involves a fresh install and a guaranteed recovery if
even that goes wrong. Consider waiting so you can time the upgrade in
to when you replace the hardware then you can migrate with zero risk.
Have fun,
John
"Ron Wolff" <ro...@scas.com> wrote in message news:<ub923j...@corp.supernews.com>...
To be fair to SCO, almost every SCO person (tech types) I ever talked to
about upgrades said DON'T...
> Cancel the weekend (pretend your Aunt died, anything). Then come up
> with a plan that involves a fresh install and a guaranteed recovery if
> even that goes wrong.
I don't know his hardware configuration, but for clients that are really
paranoid about downtime, we just sell them a new root hard drive. That way,
if things go bad during the reinstall (or even after a few hours of running
live after the reinstall), we can just swap the old one back in and they'll
be up and running in minutes.
Fabio
I've done in place upgrades and re-installs.
However before doing anything on a system any more I umount all
except / [and I have always made that only about 1GB and kept all
else out of it].
Then I make at least two verified backups of those tapes. Make
sure there are at least two sets of BackupEdge/Lone-Tar recovery
disks, and have a go.
Then if it fails reboot with the recovery disks and >remake the
file systems< . That is the only way to make sure you are back in
the state from which you started. Unless you have hot-swap drives
this will be easier than Fabio's suggestion.
I've had vendor supplied drivers corrupt a system during an
install and then having the drivers remove more than just
themselves when removing them. A bad experience a long time ago
and an extra 10 minutes to back up just / is awfully cheap
insurance. That's another good reason for not having everying
under /
Using JPR's script and a complete reinstall you do have a
benchmark/checkpoint as to the system. If you didn't do all the
work on the system up to this point, then a reinstall is the only
way you can be sure that the system meets your specifications.
Bill
--
Bill Vermillion - bv @ wjv . com