Bug 26468 - Unable to upgrade (6.2 -> 7.0), RAID
Summary: Unable to upgrade (6.2 -> 7.0), RAID
Status: CLOSED NOTABUG
Alias: None
Product: Red Hat Linux
Classification: Retired
Component: anaconda
Version: 7.0
Hardware: i386
OS: Linux
medium
high
Target Milestone: ---
Assignee: Erik Troan
QA Contact: Brock Organ
URL:
Whiteboard:
Keywords:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2001-02-07 16:01 UTC by Rupa Schomaker
Modified: 2007-04-18 16:31 UTC (History)
0 users

(edit)
Clone Of:
(edit)
Last Closed: 2001-02-08 20:12:08 UTC


Attachments (Terms of Use)

Description Rupa Schomaker 2001-02-07 16:01:42 UTC
From Bugzilla Helper:
User-Agent: Mozilla/4.0 (compatible; MSIE 5.5; Windows NT 5.0)


Custom boot disk (ATA100 support) + RH7 CD.  The system is running RAID1 
on all partitions.  If I select the upgrade path, I get the following 
exception.  I cannot seem to get it to save to floppy (i put in a new 
floppy and the disk moves and then I'm asked for another one, rinse and 
repeat) so I'm typing it in.  Hopefully I'll have gotten everything.  

The RAID is autostarted (all are fd and the status of the raid starting 
goes by at boot).  Note: that I must manually create each raid device 
in /dev, and do a raidstop on each too boot cleanly.  Simply ctrl-alt-del 
causes an oops.

=========

Traceback (inntermost last):
  File "/usr/bin/anaconda", line 438, in ? intf.run(todo, test = test)
  File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/text.py", line 1028, in 
run
    rc = apply (step[1](), (dir,) + step[2])
  File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/text.py", line 245, in 
__call__
    parts = todo.upgradeFindRoot ()
  File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/todo.py", line 841, in 
upgradeFindRoot
    mdList = raid.startAllRaid(drives)
  File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/raid.py", line 63, in 
startAllRaid
    isys.raidstart(devName, deviceList[0])
  File "isys.py", line 38, in raidstart
    _isys.raidstart(fd, "/tmp/member")
SystemError: (17, 'File exists')

======

/tmp/member does exist -- looks like a big binary file.

Resolution: stop all raid devices *before* starting the upgrade thing.  If 
I go to the bash prompt (VC2), do a:

cd /dev; for i in 1 2 3 4 5 6 7 8 9 10; do mknod md$i; done
for i in md*; do raidstop /dev/$i; done

Then let the upgrade continue.

Reproducible: Always
Steps to Reproduce:
1. Do an upgrade install
2.
3.

Comment 1 Erik Troan 2001-02-08 20:12:03 UTC
Can you try this again, and 'cat /proc/mdstat' when you get the traceback?
Seeing your /etc/raidtab would be good as well.

Comment 2 Michael Fulbright 2001-03-05 22:46:17 UTC
Closing due to inactivity.


Note You need to log in before you can comment on or make changes to this bug.