Re: install hangs inside of virtual machine

On Wed, Aug 7, 2013 at 2:19 AM, Thomas Fehr <[email protected]> wrote:
> On Tue, Aug 06, Lonni J Friedman wrote:
>
>> Greetings,
>> I'm trying to run an autoyast based installation of SLES11-SP3 inside
>> of a KVM based virtual machine (host OS is Fedora19-x86_64, if that
>> matters), and it is consistently hanging at:
>> Sending dhcp request to eth0
>>
>> It sounds quite similar to this bug:
>> https://bugzilla.novell.com/show_bug.cgi?id=799496
>>
>> however, I'm not installing over NFS at all.  I'm invoking the
>> installation from local media, and pulling down all components via
>> http.  This is the isolinux 'append' entry that I'm using:
>>
>> textmode=1 initrd=sles113_64bit.img
>> autoyast=http://172.20.232.152/distros/suse/os/SLES11-SP3/x86_64/autoyastcuda.xml
>> install=http://172.20.232.152/distros/suse/os/SLES11-SP3/x86_64
>
> The problem you are experiencing happens before yast gets started.
> It happens during setup of installation environment by linuxrc.
>
> If you use a http-URL for installation, linuxrc needs to set up the
> network to be able to access installation. Therefore it sends a
> dhcp request for eth0. Why this hangs, I cannot answer. Maybe your
> virtual machine is not set up for network connection to your DHCP
> server.

I'm quite certain that the problem isn't an inability to connect to
the DHCP server.  I'm using the same virtual machine configuration for
many different types of Linux distributions (and even Windows), and
the only one that fails at all is SLES11-SP3.  Even if I had some how
misconfigured the VM to not allow DHCP traffic, I'd expect a sane
failure message, not a complete & indefinite hang.  Shouldn't the
attempt to obtain a DHCP lease timeout at some point if that was the
actual problem?  That's certainly how I've seen all other OS's fail
when there was no response from a DHCP server.
-- 
To unsubscribe, e-mail: [email protected]
To contact the owner, e-mail: [email protected]