Bug 199459

Summary: doesn't write out partition tables
Product: [Fedora] Fedora Reporter: Bill Nottingham <notting>
Component: anacondaAssignee: Anaconda Maintenance Team <anaconda-maint-list>
Status: CLOSED RAWHIDE QA Contact: Mike McLean <mikem>
Severity: high Docs Contact:
Priority: medium    
Version: rawhideCC: rvokal, wtogami
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: 2006-07-20 22:12:46 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 Bill Nottingham 2006-07-19 17:53:33 UTC
anaconda-11.1.0.60-1.i386.rpme:

I have a RHEL 4 install currently:

 /dev/hda1 /boot  ext3
 /dev/hda2 /      ext3
 <large chunk of free space>

I choose 'custom'. I create a new 2G primary partition /dev/hda3 for '/', format
as ext3.

I choose /dev/hda1 as /boot, choose 'do not format'.

When it comes time to format filesystems, I get:

 "An error occured trying to format /dev/hda3".

If I look at the drive, /dev/hda3 was never added to the partition table.

Comment 1 Jeremy Katz 2006-07-19 19:03:00 UTC
Any kernel messages about being unable to write it?  nothing here has really
changed...

Comment 2 Bill Nottingham 2006-07-19 19:04:48 UTC
Nope. From a brief look over tty2-4, it almost appears that it didn't even try.
What steps is it supposed to write?

Comment 3 Jeremy Katz 2006-07-19 19:23:54 UTC
The partition table should get saved in the turnOnFilesystems step -- the only
reason it wouldn't is if parted failed to write it out for some reason (or the
kernel failed to notice the write)

Comment 4 Bill Nottingham 2006-07-19 19:57:44 UTC
tty3 is...
moving (1) to step confirminstall
moving (1) to step install
moving (1) to step enablefilesystems
formatting / as ext3
Format command: ['/usr/sbin/mke2fs', '/tmp/hda3', '-i', '4096', '-j' ]
tty5 is...
mke2fs: No such device or address while trying to determine filesystem size

There is a lockdep warning about anaconda attempting to acquire
(&bdev->bd_mutex) while already holding the lock.

No other useful information, really.

Comment 5 Jeremy Katz 2006-07-19 20:04:09 UTC
I blame the kernel ;-)

What are you doing using regular IDE anyway? :P

Comment 6 Bill Nottingham 2006-07-19 20:08:20 UTC
Because the other testbox is ahci/ata_piix, which works even worse at the moment? :P



Comment 7 Jeremy Katz 2006-07-20 22:12:46 UTC
Actually, after more investigation, this is Peter's fault ;-)

The mpath startup code was calling into refreshDevices which then resets
everything to its "on-disk" state.  I've stopped it from doing that for now
which hopefully won't break mpath/dmraid