Red Hat Bugzilla – Bug 56198
uunhandled exception, installer fails
Last modified: 2007-04-18 12:38:16 EDT
From Bugzilla Helper:
User-Agent: Mozilla/4.78 [en] (X11; U; SunOS 5.6 sun4u)
Description of problem:
tryed the basic workstation install from installer, grub, and gnome
I have included the write to floppy file from the installer
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Tryed reinstalling 7.2 several times
2. Three times I received unhandled exception. I finally wrote file out to
3. The other times I did not get the unhandled exception, the final reboot
failed after getting successful installation.
Intel d815eeaal mthbrd,
PIII, 733, 256 cache, 133fsb, FCPGA processor
Riva tnt-2 AGP video card
Dell 1710 17" monitor
40 gig IDE H.D.
384 MB memory
plextor px-w1610a/sw cd-rw drive
Could you attach the dump of the exception that you recieve?
I have submitted the dump and was wondering if any progress has been made. I
see this bug is still listed as needing info.. Let me know I am really waiting
for a result so i can install this ver on my computer...
I also have the same problem...please help
It has been 2wks since original bug submission. I have heard nothing. Is anyone
trying to reslove this problem? I hope to get some response. Please help.
I do not see any attachments to this bug - please try to attach the exception again.
Created attachment 39588 [details]
this is dump attachment for bug 56198
I get a very similar error, except in my case /usr/sbin/usrmod can't be run. In addition
there is one thing that does not appear on the anaconda log: a message pops up
shortly after the last package is loaded which says something like 'can't load libslang.so.1'.
I don't know the exact error, because it gets overwritten, but I am sure about the filename.
Will attach dump and hopefully more detail about the error message if I get it.
Created attachment 40087 [details]
my anaconda dump with "RuntimeError: /usr/sbin/usermod can not be run"
I know why my install keeps failing, but I don't know how to fix it.
anaconda fails because it can't run /usr/sbin/usermod, because
/usr/lib/libslang.so.1 can't be loaded, because it isn't there.
I have three attachments from my latest attempt: problog.txt , the detailed account;
anacdump2.txt, the latest anaconda dump; and usrlib.txt which is an ls -l of the
Created attachment 40107 [details]
more detailed log of the problem & findings
Created attachment 40108 [details]
another anacdump.txt from 3rd attempt
Created attachment 40109 [details]
ls -l /usr/lib from failed install, to show missing libslang.so.1
On closer examination I find that there isn't a /usr/sbin/usermod on the disk either, and I guess maybe it runs off the ramdisk or something. So maybe libslang.so.1 is
supposed to load from the ramdisk. There is a /usr/sbin/kudzu which contains the string ' libslang.so.1'.
I'm going to stop for now; I'm also providing a ls -AlR of the whole install.
Created attachment 40127 [details]
ls -AlR of the root partition after failed install
OK. have some more info. Now it looks like anaconda
can't run '/usr/sbin/usermod' because there is no 'usermod'
anywhere on the system, even within the 'install' root.
Also, there is no 'chfn' in /usr/sbin or anywhere else
- this is the command which fails for email@example.com
Actually there is /mnt/sysimage/usr/share/doc/nss_ldap-172/pam.d/chfn
but that is not executable.
Attachment INSROOT.TXT is a ls -lA of the 'install' root
and ls -lRA of the following subdirectories:
This info was obtained by going into 'debug' after the failed
install, and getting a shell by running "os.system('sh')"
An 'ls -lAR' of what is '/mnt/sysimage' was previously provided
as attachment 40127 [details].
There is a '/usr/lib/libslang.so.1' which is
However, there is no /usr/sbin/usermod in this
environment or any 'usermod' file that I can find
Created attachment 40227 [details]
list of files in the 'install' root and /mnt/runtime after faied install
I tried the install using 'boot.img' floppy, from a set of .iso files contained on
a hard disk (rather than the FTP install done previously) and it worked
fine this time. So I'm OK now. I have kept one of the failed installs on
a partition in case there is anything that might be useful to you there.
I get the same error as the anacdump.txt file dated 12-04-2001 all the way up
to the error at line 208. After that I get
File "/tmp/updates/packages.py". line 793, in doPostinstallf.close()
IOError: [Errno: 28] no space left on device.
I can not dump the file because at that point my floppy drive fails. Is there
a special way to dump it to the floppy, ie I need a formatted disk or something?
If I load the workstation version it works. If I do custom and select
everything, I get the error. I let the Red Hat software partition my drive and
the drive is 36GBytes. I just purchased Red Hat and this is my first load.
The error occurs because of a bad CD. By using the hard drive install you got
around the issue.
There was no CD involved. Install was from FTP image derived from downloaded ISO image (and MD5 checksum was OK). There may have been a problem with
W32-based FTP server. The hard drive install , which worked, was from the same ISO images used to build the FTP directory from which the failed install occurred.