Lost Partitions on Xenix..
JJ
know..):
1) Removed an existing primary/only drive from a xenix machine with
intention of copying data to a Unix machine (already existing with multiple
drives)
2) Unplugged existing secondary master IDE drive from Openserver 5.0.4
machine, which had cyls/head/sectors set in bios (not auto)
3) Plug in xenix drive as secondary master on IDE chain
4) Run 'mkdev hd' - add IDE drive - added device /dev/rdsk/4s0, showed
existing XENIX partition, and reported 'hard disk initialization complete'
5) Reboot Unix server, invoke mkdev hd again - Instead of invoking divvy, it
added new device /dev/rdsk/5s0, no drive paramters (since there was no drive
for this...)
6) Realized that divvy wasn't going to be invoked by mkdev hd, we attempted
to manually run divvy on /dev/rdsk/4s0, it said a valid division table was
not found.
7) We realized at this point the drive paramaters were that of the old
(original) Secondary Master IDE (a 7GB Drive), and therefore changed them in
both the bios and using dparam to the correct paramaters for this 204MB IDE
drive
8) Still unable to run divvy on the drive, and now we are out of options...
Any help or advice on this? It is over at data recovery labs now, but I am
afraid they may come back and say we are screwed...and I believe you all may
know for certain if we are...
Help?
Dave Redit
Steve Fabac
>
> Here's what happened, (please don't tell me what we did wrong here...we
> know..):
>
> 1) Removed an existing primary/only drive from a xenix machine with
> intention of copying data to a Unix machine (already existing with multiple
> drives)
I have done this many times when upgrading a client from Xenix to SCO
UNIX. I have a 1G IDE Maxtor with Openserver 5.0.0 Enterprise on one
partition, SCO Xenix 2.3.4 on another, and 32M Dos on a third.
The Openserver 5.0.0 is configured with five hard drives hd00, hd10,
hd20, hd30, and hd40. These are various combinations of IDE and SCSI
(on AHA-1542 and AHA-2940) so that I can mount IDE or SCSI Xenix
disks and the new system's SCSI disk and transfer files.
>
> 2) Unplugged existing secondary master IDE drive from Openserver 5.0.4
> machine, which had cyls/head/sectors set in bios (not auto)
>
> 3) Plug in xenix drive as secondary master on IDE chain
Bet your 5.0.4 system was set for LBA to support large IDE drives.
Xenix MUST be run in native (un-translated) mode. You should have
detected the IDE drives in CMOS and made sure that you chose "large"
for untranslated (non LBA).
Also, did you add Xenix file system support to you 5.0.4 system kernel?
With out that, you would not be able to get the data off the disk.
>
> 4) Run 'mkdev hd' - add IDE drive - added device /dev/rdsk/4s0, showed
> existing XENIX partition, and reported 'hard disk initialization complete'
There should have been no need to run mkdev hd. Just fdisk to see if
you can see the disk and then divvy to read the partition table.
you might have to run divvy hd1X where X is the Xenix partition as shown
by fdisk. Use divvy to name the partitions.
Openserver 5.0.0 for sure can access the Xenix divvy table. You should
only have had to name the partitions on the Xenix disk and then exit,
installing the division as shown. Note: Be sure that none of the
division are marked "yes" for new or divvy will clean them out and you
will lose all your information in the partition.
>
> 5) Reboot Unix server, invoke mkdev hd again - Instead of invoking divvy, it
> added new device /dev/rdsk/5s0, no drive paramters (since there was no drive
> for this...)
>
> 6) Realized that divvy wasn't going to be invoked by mkdev hd, we attempted
> to manually run divvy on /dev/rdsk/4s0, it said a valid division table was
> not found.
>
> 7) We realized at this point the drive paramaters were that of the old
> (original) Secondary Master IDE (a 7GB Drive), and therefore changed them in
> both the bios and using dparam to the correct paramaters for this 204MB IDE
> drive
Yikes! dparam is not needed in this case. Just use the IDE drive
detection in CMOS and choose non LBA.
>
> 8) Still unable to run divvy on the drive, and now we are out of options...
>
> Any help or advice on this? It is over at data recovery labs now, but I am
> afraid they may come back and say we are screwed...and I believe you all may
> know for certain if we are...
>
> Help?
>
> Dave Redit
I would put the drive back in the Xenix box, boot your Xenix emergency
boot floppy and type mount -r /dev/hd0root /mnt then see what is on the
disk. You may have to run the Xenix dparam to reset the cyl, heads, and
sectors back to non LBA values. also run fdisk an make sure that the
Xenix partition is marked active.
If you did not run divvy -I or divvy -m from the 5.0.4 system on the
hard disk, you probably can fix the parameters and get the disk booting
in the Xenix box again. Even if you did, you may be able to run
"divvy -c 1 -b 1 -p 0 -I" from a Xenix emergency boot floppy and re-set
the Xenix division table: DO THIS ONLY ON A COPY OF YOUR DISK. NOT
ON THE ORIGINAL XENIX DISK. FOR THIS TO BE SUCCESSFUL, YOU MUST HAVE
THE EXACT LAYOUT OF THE XENIX DISK AVAILABLE SO THAT YOU CAN ENTER THE
STARTING AND ENDING BLOCKS. BE SURE THAT YOU PREVENT CREATING NEW FILE
SYSTEMS.
Good luck.
Steve Fabac
S.M. Fabac & Associates
816/765-1670