On Thu, Feb 06, 2014 at 02:43:07PM -0500, Paul Wouters wrote:
>
> I'm using a minimal netinstall version of fedora20 for testing using
> KVM. We very often cycle these machines (once per test, we run hundreds
> of tests)
I run minimal installs all the time, with either a script like that[1]
or with 'virt-builder' invoked as:
$ virt-builder fedora-20 --format qcow2 --size 10G
$ virt-install --name f20vm1 --ram 2048 --disk \
path=/export/images/fedora-20.img --import
I'm not seeing any such issues you describe, but, of course, the above
is not a netinstall, so something could be different in your
environment.
>
> Regularly, we get tests failing because the VM does not boot within 60
> seconds, and seems to hang at:
>
> "a stop job is running for User Manager for 0"
>
> The machine is still in single user mode, so I cannot get any more
> information out of it.
Not even `journalctl` to display boot messages with error priority?
$ $ journalctl -b
$ journalctl -b -p err
> The message is rather useless, because it does
> not tell me anything about the problem. Which job is still running? What
> is it waiting for? What's a User Manager? Who is "0"?
>
> I'm not even sure at what component we are looking for to report a bug :(
>
> Does anyone have more information for me?
A couple of related bugs[2][3]. The RH bugzilla URL below has some
journalctl diagnostics information.
[1] https://github.com/kashyapc/virt-scripts/blob/master/create-guest-qcow2.bash
[2] https://bugs.freedesktop.org/show_bug.cgi?id=70593
[3] https://bugzilla.redhat.com/show_bug.cgi?id=1023820
--
/kashyap
--
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxxx
https://admin.fedoraproject.org/mailman/listinfo/devel
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
