Red Hat Bugzilla – Bug 80855
Installer crashed at partition section
Last modified: 2007-04-18 12:49:25 EDT
Description of problem: Using 'linux ks=floppy' on boot to install RH 7.3. Attached is the kickstart file I am using (personal data removed; Ok, just realized Python dump inserts data in file :-/). Stepped through GUI screen (problem occurs with text mode also). Crashed happens after mouse options and before partition setup. When I click the Debug button, "KeyError: hda1" is displayed. System had same installation before hand. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Boot from CD #1. 2. At boot prompt, type 'linux ks=floppy'. Floppy should have ks.cfg. 3. Click Next button in GUI. 4. Watch installer crash... Actual Results: System requires reboot. Expected Results: Haven't performed a KS install before. I would think Disk Druid would appear with my partioning information. Additional info: Since output was sent to console, I cannot add the error messages displayed.
Created attachment 89021 [details] Kickstart config This is the KS config I was using. Personal info has been removed.
Created attachment 89022 [details] Anaconda dump file
There are no partition tables on the drives according to the kernel partition scan, and you have 'zerombr yes' commented out. You need to enable this line so it can put a clean (empty) partition table on drives w/o them.
Yes, I originally used 'zerombr yes', which produce the initial error. Afterwards, I did change some things around to see if I could get it to work. When I used 'zerombr yes', I had the following: clearpart --all part raid.1 --onpart hda1 --badblocks part raid.7 --onpart hdb1 --badblocks part raid.2 --onpart hda2 --badblocks part raid.8 --onpart hdb2 --badblocks part raid.3 --onpart hda3 --badblocks part raid.9 --onpart hdb3 --badblocks part raid.5 --onpart hda5 --badblocks part raid.11 --onpart hdb5 --badblocks part raid.6 --onpart hda6 --badblocks part raid.12 --onpart hdb6 --badblocks raid / --fstype ext3 --level=RAID1 raid.2 raid.8 raid /boot --fstype ext3 --level=RAID1 raid.1 raid.7 raid swap --fstype swap --level=RAID0 raid.5 raid.11 raid /home --fstype ext3 --level=RAID1 raid.6 raid.12 raid /var --fstype ext3 --level=RAID1 raid.3 raid.9 Other than 'zerombr' and 'clearpart --all', the rest was produced by Anaconda.
This should be handled somewhat better in 8.0 and even a little bit better in Phoebe.