From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5)
Description of problem:
Doing an FTP install after booting from CD-ROM anaconda barfs after
successfully loading stage2, complaining it can't find libraries.
From the Python backtrace:
Could not find platform independent libraries <prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
'import site' failed: use -v for traceback
Traceback (most recent call last):
File "/usr/bin/anaconda", line 28, in ?
import sys, os
ImportError: No module named os
So, it seems like some sort of PATH problem.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. burn boot.iso to CD, start install
2. load stage2 image from FTP site and continue
3. install barfs
Actual Results: Installer barfs
There have been a lot of fixes in the past few days so this might be
better now. Assigning to Chris to double check.
I'm unable to duplicate this problem, so I'll assume one of the million changes
we've made in the last couple days fixed it. Feel free to reopen if you see it
in later rawhides or if we re-break things.
I keep seeing this problem, also with the latest boot.iso (today) and older ones
(yesterday). I verified that the md5sums on the ISO and the CD are the same (in
fact, I had k3b verify that). What should I do to effectively debug this?
I'm not able to reproduce this problem inside VMware or on a Dell Optiplex
GX260, but I can reproduce it repeatedly and reliably on Dell Optiplex GX110
machines (tested 5 so far).
I'm still seeing this problem with the latest and greatest version of 5 minutes
ago (and various versions before that). I can't switch to another console to
debug the problem when it occurs, because the barfing of the installation
process halts the operating system. What should I do to debug this? Patching
anaconda and rebuilding the whole distribution is not a problem for me (done
I just don't know where to debug :S
Still can't reproduce this one. What's the last screen you're seeing before it
fails? Can you reproduce this using a method other than an FTP install? Also
can you test your memory using memtest86?
It downloads the image fine from FTP, then it barfs after it says "starting
anaconda". I'm pretty sure it's *not* a memory problem, because I can reproduce
this on at least 5 identical boxes (all Dell Optiplex GX 110 machines), but not
on any others. Which install method would you recommend (there are no ISOs
available, are there)?
I got the same problem on my system. Fedora-Devel (version of 03/13)
inside VMWare 5.0b3 (on the top of FC3).
In my case it happens when doing network install (from boot.iso) and
specifying "linux text" to perform text install.
Then perform HTTP install.
Apparently when this happens there are no python modules in netsg2.img, at least
according to gb and msw (they're pretty reliable sources)
Created attachment 112155 [details]
rawhide March 19th 2005 - it happened again
Note that this is a text install only failure, while graphical install may work.
This is because...
<msw> warren: it didn't have any python modules in netstg2.img or hdstg2.img
<warren> msw, oddly it works for graphical install, but not text
<msw> warren: right
<msw> warren: stage2.img has all the modules you need
<msw> warren: but if you do a ftp,http,hd install in text mode, you'll use the
I find this really odd, because it did work when I used the same CD image and
same ftp install from another machine (other hardware) and from inside VMware,
but failed on all Dell GX-110 machines I have. I'm using graphical installs btw.
We might be looking at two different problems.
Okay, I believe this is being caused by errors in building the netstg2.img.
Will roll a new anaconda package tonight. Please test a rawhide tree from
tomorrow or later and verify that it works for you.
Doesn't fix it for me. Is there some debug version I could run?
Ah, the error is still occurring because I only fixed half the problem. Sorry
about that. Please try tomorrow or later.
Having same problem where can I get the new netstg2.img to test? (Once it's done)
Copied stage2.img to netstg2.img on http server as implied above. http install
(booted from cd, "linux text askmethod") works now.
I don't get this problem anymore but a bunch of other errors unrelated to this
one instead ;-) So, I think this one can be safely closed...
Closing based on bug submitter comments.
*** Bug 152610 has been marked as a duplicate of this bug. ***
I tried this workaround (copying the stage2.img over the netstg2.img) and it
didn't work for me. I'm sync'd to the latest dev tree as of last night
(2005-04-01) which is what I'd been told in #fedora-devel by a number of people.
I'm not sure what I should try at this point...