Red Hat Bugzilla – Bug 433658
Installer / rescue from 2008-02-19 not handling bad partitions usefully
Last modified: 2008-03-10 13:23:47 EDT
Description of problem:
I have been trying to install rawhide with encryption on top of raid and have
been having some problems.
I was trying out the boot.iso from 2008-02-19 and it crashed after confirming
that I really wanted to install rawhide. I think this was because it couldn't
figure out what to do with partitions that were incompletely set up when I tried
using the alpha iso to do the install.
Then I tried booting in rescue mode so that I could use badblocks to overwrite
the partition table (I had to do this when using the alpha iso) and got prompted
for a password to the partitions (which shouldn't have been encrypted) and the
raid arrays. The pasword didn't seem to work. After skipping ahead (since I
didn't need to mount the filesystems) anaconda crashed without the option to
The error messgae was:
rootparts = ananconda.id.diskset.findExistingRootPartitions(upgradeany =
File "/usr/lib/anaconda/partedUtils.py", line 775, in findExisting
except SystemError, (errno, msg):
ValueError: need more than 1
value to unpack
install exited abnormally [1/1]
I am attaching the saved traceback from the install attempt (not the rescue
Version-Release number of selected component (if applicable):
I am not sure of the version built into the boot.iso.
100% with the current hard drive contents, reproducing that might prove tricky.
Steps to Reproduce:
1. Use alpha.iso to try to install with encryption on top of raid.
2. Use boot.iso do to an install after the above step has failed.
Created attachment 295436 [details]
Traceback from install attempt
Created attachment 296121 [details]
Anaconda traceback dump
Today's images have the same problem. After partially installing with
encryption over raid, further installs were blocked until the partition
table was wiped.
*** Bug 435210 has been marked as a duplicate of this bug. ***
The traceback won't happen as of tomorrow's rawhide, but there might still be
problems lingering. So I'll leave this open to be tested then.
This bug has been fixed in current anaconda-126.96.36.199