Red Hat Bugzilla – Bug 100802
Anaconda barfs when it's about to format ths disks
Last modified: 2007-04-18 12:56:07 EDT
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.0.2)
Description of problem:
The dump is attached. This occurs in the graphical RH setup just
after package selection, when we are about to format the disks
Nonclassical things I do/have/did:
- Two IDE drives /dev/hde and /dev/hdg in software RAID-1 configuration
- Boot device is going to be /dev/md1 NOT /dev/md0 (is this a sin?),
- Tried to modify the root password after having set it (with 'back'
and that didn't work, incidentally)
- Drink lots of green tea and eat bananas while waiting for
installer to finish
Steps to Reproduce:
1. Set up RH 9
2. Get to after the point where package info was entered
Actual Results: Bad things
Expected Results: Good things
Created attachment 93154 [details]
Anaconda dump file
Uh oh! Something must have gone wrong with the assignation of the md
I restarted the setup process and there are md devices
md3 md7 md8 md5 md4 md1 md6 md1 md0 md2
I was sure I had them set to
md1 md2 md3 md5 md6 md7 md8 md9 md10 md11 md12 md13
The idea was to reflect the structure of the underlying hde/hdg partitions.
I will be reasonable now and count from 0 and not do jumps in the md series.
Did you have preexising RAID devices on the system before you reinstalled?
The system had a RH 8.0 installation but the two disks were not in a RAID
setup. I decided to scratch everything, but keep the partitioning of the
disks (it was exactly alike on both disks). So I just marked them as
'software raid' then paired them into mirrors in disk druid:
md1 = hde1 + hdg1 (primary, will hold /boot)
md2 = hde2 + hdg2 (primary, will hold /)
md3 = hde3 + hdg3 (primary, will hold <swap>)
md5 = hde5 + hdg5 (secondary, will hold /tmp)
md6 = hde6 + hdg6 (secondary, will hold /home)
md7 = hde7 + hdg7 ...etc...
md8 = hde8 + hdg8
md9 = hde8 + hdg9
md10 = hde10 + hdg10
md11 = hde11 + hdg11
md12 = hde12 + hdg12
md13 = hde13 + hdg13
This looks like it should be working better in our current releases.