Bug 159079

Summary:	anaconda crash "md device not found" on /dev/i2o devices
Product:	[Fedora] Fedora	Reporter:	Warren Togami <wtogami>
Component:	anaconda	Assignee:	Chris Lumens <clumens>
Status:	CLOSED RAWHIDE	QA Contact:	Mike McLean <mikem>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	rawhide
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:	2005-08-15 16:39:19 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:	150221

Description Warren Togami 2005-05-29 00:28:46 UTC

Anaconda crashed while attempting an install of rawhide 20050528 on software
RAID using two I2O arrays /dev/i2o/hda and /dev/i2o/hdb as disks.

http://people.redhat.com/wtogami/temp/anacdump-i2o-md.txt

# partition table of /dev/i2o/hda (identical on /dev/i2o/hdb)
# (hda1 -> md1) /boot
# (hda2 -> md2) swap
# (hda3 -> md0) /

unit: sectors

/dev/i2o/hda1 : start=       63, size=   514017, Id=fd, bootable
/dev/i2o/hda2 : start=   514080, size=  1012095, Id=fd
/dev/i2o/hda3 : start=  1526175, size= 70236180, Id=fd
/dev/i2o/hda4 : start=        0, size=        0, Id= 0

Comment 1 Ronny Buchmann 2005-05-29 23:07:29 UTC

I have the same crash with existing arrays (build from regular ide disk partitions).

partedUtils.DiskSet.mdList is empty for here

I solved that by starting the raids in fsset.py:mdadmConf().

RCS file: /usr/local/CVS/anaconda/fsset.py,v
retrieving revision 1.253.2.1
diff -u -p -u -r1.253.2.1 fsset.py
--- fsset.py    25 May 2005 18:53:39 -0000      1.253.2.1
+++ fsset.py    29 May 2005 22:56:12 -0000
@@ -1123,6 +1123,9 @@ class FileSystemSet:
     def mdadmConf(self):
         raident = 0

+        diskset = partedUtils.DiskSet()
+        diskset.startAllRaid()
+
         cf = """
 # mdadm.conf written out by anaconda
 DEVICE partitions
@@ -1135,6 +1138,7 @@ MAILADDR root
             raident +=1
             cf = cf + ent.device.mdadmLine()

+        diskset.stopAllRaid()
         if raident > 0:
             return cf
         return

I assume there may be a better way to fix that.
I think that should be targeted at FC4.

Comment 2 Chris Lumens 2005-05-31 17:18:32 UTC

Ronny - can you explain in more detail what you have for partitioning before
starting?

Comment 3 Chris Lumens 2005-05-31 21:02:00 UTC

Applied a fix for Ronny's problem.  Warren, please test with FC4 at your leisure
and advise if it works for you or not.

Comment 4 Rahul Sundaram 2005-08-15 16:39:19 UTC


Warren, 

Kindly reopen if this one is not fixed yet