Red Hat Bugzilla – Bug 170530
'Upgrade' on a RAID 1 system complains of duplicate drive identifiers.
Last modified: 2008-08-02 19:40:34 EDT
Description of problem:
'Upgrade' on a RAID 1 system complains of duplicate drive identifiers. New
installs & full a re-install over existing FC4 both work OK without this error.
Version-Release number of selected component (if applicable):
Anaconda as shipped in: 2.6.11-1.1369_FC4
Steps to Reproduce:
1. Start a installation and select the Upgrade option
I am advised that two drives have the identifier "LINUX" and can then only
reboot. Cannot recall the exact words of the error but this is the implication.
Should go ahead and upgrade existing installation - replacing damaged files.
Did you receive a traceback (the "Exception Occured" dialog would pop up with
several options) or just an error message? If you got a traceback, please
attach it to this bug report.
No additional info at all. Just the bald statement about duplicate drive
identifiers. When I click OK (I think) all I got was the option to reboot.
I didn't bother to retry, I just did a new install, loading everything again
over what was already there. No error message doing that or when I loaded FC4
for the first time. The first install was only a few days earlier as I'm
totally new to Linux and the relaod was because something convinced my that I
had a corruption somewhere and this seemd the quickest way to clear it.
Hope this helps.
I saw something similar to this when trying to upgrade a RedHat 9 system which was installed on a
Promise RAID (set in the BIOS). The installer complained that two / partitions were found. This may also be
related to a bug I noticed in upgrading a Fedora Core 2 system set up with software RAID-1 to FC4. In that
case, I found that only one of the two RAID-1 slices survived the upgrade and I had to reattach and resync
the others. Finally I noticed yesterday that when clean installing on a set of newly created software RAID-1
partitions that only one slice of each RAID-1 partition was up todate after the installation and that the
other slices were resync'd to them. So it seems that the Fedora Core installer doesn't really write to both
slices of a RAID-1 partition under certain circumstances.
This report targets the FC3 or FC4 products, which have now been EOL'd.
Could you please check that it still applies to a current Fedora release, and
either update the target product or close it ?