Bug 101054

Summary: Kickstart using md raid devices fails
Product: Red Hat Enterprise Linux 2.1 Reporter: Tim Perry <tperry>
Component: anacondaAssignee: Michael Fulbright <msf>
Status: CLOSED RAWHIDE QA Contact: Mike McLean <mikem>
Severity: medium Docs Contact:
Priority: medium    
Version: 2.1   
Target Milestone: ---   
Target Release: ---   
Hardware: i386   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2003-07-28 22:14:03 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:

Description Tim Perry 2003-07-28 20:14:18 UTC
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.2.1) Gecko/20021130

Description of problem:
This is on a RH Enterprise WS 2.1 Install.

According to the documentation in order to create a Raid partition, you specify
the device md0 md1 in the raid section of the kickstart file: 

ie:

part raid.01 --size 256 --grow --asprimary --ondisk hda
part raid.02 --size 256 --grow --asprimary --ondisk hdc
part raid.03 --size 2048 --asprimary --ondisk hda
part raid.04 --size 2048 --asprimary --ondisk hdc
raid / --level 1 --fstype ext3 --device md0 raid.01 raid.02
raid swap --level 1 --device md1 raid.03 raid.04

However when doing this I have found one of two things happen, If I only specify
on md device, (ie I get rid of swap) then the system installs, but fails on
reboot.  After some investigation it appears that the installer mislabels the md
device to /dev/mdmd01 instead of /dev/md01 in the /etc/raidtab and
/etc/grub.conf.  After fixing these two files the system works fine.

However, if I include two raid devices it fails during the install, unable to
format /dev/mdmd0.

Workaround:
-----------
In the kickstart file do not use md in the device name, I figure that somewhere
in the install code, someone is automaticly appending /dev/md to this string. 
When I just put --device 0 instead of --device md0 it seems to work as in the
example below

part raid.01 --size 256 --grow --asprimary --ondisk hda
part raid.02 --size 256 --grow --asprimary --ondisk hdc
part raid.03 --size 2048 --asprimary --ondisk hda
part raid.04 --size 2048 --asprimary --ondisk hdc
raid / --level 1 --fstype ext3 --device 0 raid.01 raid.02
raid swap --level 1 --device 1 raid.03 raid.04

If this is the intended way, it is a bug with the documentation not the code. 
Either way it needs to be fixed one place or the other.


Version-Release number of selected component (if applicable):


How reproducible:
Always

Steps to Reproduce:
1.put the above partition info in a network kickstart
2.Kickstart the server
3.it fails (either on reboot or during install depending on the described way of
implementation)
    

Actual Results:  Failure to reboot, /dev/mdmd0 is in /etc/grub.conf and /etc/raidtab


Expected Results:  System should have come up cleanly

Additional info:

Comment 1 Jeremy Katz 2003-07-28 22:14:03 UTC
In 2.1, it was supposed to be '0', '1', etc.  Current code allows many more
variations.