Re: [KVM_AUTOTEST] unattended installs take 2

[Lucas Meneghel Rodrigues <lmr@redhat.com>]

On Mon, 2009-06-22 at 02:20 -0400, Michael Goldish wrote:
> ----- "Lucas Meneghel Rodrigues" <lmr@redhat.com> wrote:
> 
> > On Thu, 2009-06-18 at 17:50 -0400, David Huff wrote:
> > > Second pass at the unattended install test.  Both Linux and Windows
> > guests are
> > > working however currently it just uses the existing boot test and
> > extends
> > > timeouts.  We still need a good way of determining if an unattended
> > install
> > > completed without error. For Linux guest we should be able to run
> > something
> > > similar to a regular boot test, after reboot try to ssh in.  For
> > windows guest
> > > we still have to run setup in order to have shh up.
> > > 
> > > Requires the processor patch as well as the patch to "strip and
> > split" patches.
> > > 
> > > Scripts still uses loop back mounts for now we can upgrade this in
> > the future
> > > however most generic way of creating and manipulating disk images.
> > > 
> > > 5 patches in this set
> > > 0005-Modified-boot-test-in-kvm_test.py.patch
> > > 0004-Added-two-sample-unattended-config-files-Fedora-and.patch
> > > 0003-added-unattended.sh-script.patch
> > > 0002-modified-config-file-to-run-unattended-install.patch
> > > 0001-Added-floppy-and-tftp-options-to-qemu-command.patch
> > 
> > I've been trough the changes, thank you for your work David:
> > Comments/questions:
> > 
> >  * Any particular reason why you guys wrote the PXE boot setup as a
> > shell script instead of a python module (that could be also used as a
> > stand alone program)? 
> >  * The script had some unused variables, and some extra debugging
> > statements would be helpful. I've modified the original script a bit
> > and
> > I am attaching it to this message, please verify if the changes make
> > sense.
> > 
> > Your changes are an excellent starting point for getting unattended
> > installs integrated. So things I'd like to explore:
> > 
> >  * Turn the script into a python module
> >  * See whether the development version of cobbler already handles
> > windows appropriately. Maybe for now cobbler is not an option, but we
> > can revisit it in the future.
> >  * Create an install test that would take as the parameter the type
> > of installation we want to perform (steps/unattended).
> 
> I don't see a very good reason to implement that last item, since:
> - we can already easily select tests in the config file with 'only install' or 'only unattended_install' (we can rename 'install' to 'steps_install'), so there's no advantage to adding an additional install type parameter,
> - we might want, and probably would want, to run both 'install' and 'unattended_install' in long test sets (weekly or longer ones),
> - both install tests are already quite complex -- creating a wrapper for them would make them slightly more complex
> 
> What do you think?

Fair enough, at first I thought it'd be good for the purposes of
organization, but you made a good point against it. Let's work on the
other points.

> 
> Thanks,
> Michael
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html