Bug 495552

Summary: incorrect mdadm.conf syntax prevent reboot after installation
Product: [Fedora] Fedora Reporter: Matthieu Araman <matzilla>
Component: anacondaAssignee: David Lehman <dlehman>
Status: CLOSED RAWHIDE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: urgent Docs Contact:
Priority: low    
Version: rawhideCC: pjones, rmaximo, vanmeeuwen+fedora
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2009-04-26 11:58:32 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 495581    
Attachments:
Description Flags
anaconda mdadm syntax correction none

Description Matthieu Araman 2009-04-13 19:31:35 UTC 
Description of problem:
installing Rawhide with LVM+Raid setup (see parent bug)

Version-Release number of selected component (if applicable):
boot.iso 20090404

How reproducible:


Steps to Reproduce:
1.install rawhide
2.reboot
3.
  
Actual results:
grub OK
load initrd
some hardware detected
latest message about raid + usb 
then nothing
have to ctrl-alt-suppr

Expected results:
system boot after install

Additional info:
I did some debug and think I found the problem (though once corrected, I'm bitten by another bug which corrupt the system)

with F11 beta, I boot in rescue mode
load partition
then look at /etc/mdadm.conf in the sysimage and /

obvious error -> there's a missing \n between the two lines (as I have both /dev/md0 and /dev/md1, I have two lines for raid)

runs mdadm --examine --scan, which should give the same in my understanding

same :
name of md devices
num-devices
UUID

difference :
mdadm created by anaconda -> level=1
mdadm result -> level=raid1


I edited grub.conf and did the following tests 

standard installation setup -> KO
mkinitrd redone from rescue setup (after chroot /mnt/sysimage) -> KO
mkinitrd redone + correcting the missing \n between the two lines -> KO
mkinitrd redone + correcting both the missing \n and putting level=raid1 -> Boot, load partitions, and freeze later on the firstboot screen (will open another bug for this)

I can't test the latest anaconda because of bug 495261.
Can you correct this one if still present (the missing \n at end of line + replacing level=$n by level=raid$n )?
Comment 1 Matthieu Araman 2009-04-14 10:01:15 UTC 
Created attachment 339453 [details]
anaconda mdadm syntax correction

This patch change anaconda mdadm.conf syntax generation to match the output of mdadm --examine --scan (another solution could be to just do a "mdadm --examine --scan > mdadm.conf from the install setup as the md should be started to allow installation to occur so the output should be the same in my opinion)

you'll find original and corrected working mdadm.conf at  
http://plouic.com/fedora-matzilla/rawhide20090404/
Comment 2 David Lehman 2009-04-16 17:28:26 UTC 
This should be fixed in anaconda-11.5.0.45-1.
Comment 3 Matthieu Araman 2009-04-26 11:58:32 UTC 
I confirm this is working with boot.iso/anaconda-11.5.0.45-1

The system boot correctly after installation.
Raid is correct
/etc/mdadm.conf looks correct
cat /proc/mdstat gives :
/dev/md0 (/boot, ext3 ) : active
/dev/md1 (the rest via LVM) : rebuilding in the background after rebooting