Bug 431508

Summary: anaconda selects impossible runlevel for initdefault during kickstart
Product: [Fedora] Fedora Reporter: Jon Stanley <jonstanley>
Component: anacondaAssignee: Chris Lumens <clumens>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: low Docs Contact:
Priority: low    
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: 2008-02-05 16:05:09 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:
Attachments:
Description Flags
kickstart file none

Description Jon Stanley 2008-02-05 06:00:17 UTC 
Description of problem:

With the attached ks.cfg file (used to test kickstart support for encrypted
partitions - which works beautifully BTW), runlevel 5 was the initdefault for
the resulting system, even though no windowing system was installed.  Therefore,
you get a :id x respawning too fast, disabled for 5 minutes" message.


Version-Release number of selected component (if applicable):

anaconda-11.4.0.30-1

How reproducible:

Didn't try, presume every time :)

Steps to Reproduce:
1. Kickstart using attached file

Actual results:

Runlevel 5 selected as initdefault, which doesn't work.

Expected results:

Runlevel 3 selected as initdefault.
Comment 1 Jon Stanley 2008-02-05 06:00:17 UTC 
Created attachment 293970 [details]
kickstart file
Comment 2 Jon Stanley 2008-02-05 06:12:57 UTC 
/hmmm, just noticed invalid stuff in there, like:

xconfig --startxonboot

and the --append for the bootloader (I copied this ks.cfg from what anaconda
deposited in /root/anaconda-ks.cfg and must not have looked at it too closely).

Retrying with a more sane set of options (and --nobase to speed things up) and
will report back.  I still don't think that it should have set the runlevel to 5
with no chance of it working, but some of this could likely be chalked up to
PEBCAK. :)
Comment 3 Jon Stanley 2008-02-05 06:38:43 UTC 
Yeah, sane options got this working right.  I'm gonna leave this open, but I'm
not sure how to proceed - on one hand, anaconda did exactly what I told it to. 
On the other hand, what I told it to do was impossible to ever work.  Should
there be some sort of error checking for these cases??
Comment 4 Jeremy Katz 2008-02-05 13:29:29 UTC 
In general, we take the approach of "do what is asked for, even if it seems
nonsensical" with kickstart.  I'm not really sure what (reasonable) checking we
can do as you could be doing installs of lots of different things which would
have different meanings of the *dm used in runlevel 5.
Comment 5 Chris Lumens 2008-02-05 16:05:09 UTC 
No, there's really nothing we can do here besides log some sort of warning
message for later.  It's entirely possible to run anaconda in text mode, but
want to install and use X on the installed system.  We won't be able to tell
this until much later on, as we'll have to inspect the display mode requested
and the package selections made.  At this point, I feel that too complicated of
a scheme is just bound to failure and won't ever really do what we want anyway.

In short, kickstart makes impossible things possible.  :)