F22 Atomic networking failure in OpenStack (A regression from F21!)

Thu Jul 9 22:10:17 UTC 2015

So, in an OpenStack environment I'm trying to use, F22 Atomic can't get an IP/routing info, but F21 Atomic *can*.

The first sign that something's up is a non-trivial pause on "A start job is running for Network Manager Wait Online", followed by cloud-init failing:

[   36.377396] cloud-init[721]: 2015-07-09 20:48:49,254 - util.py[WARNING]: Route info failed: Unexpected error while running command.

The complete boot log of the F22 failure is here: http://pastie.org/private/1qikbe1hu328ajszohq5a

What's interesting is that F21 Atomic comes up fine, in a completely identical environment. A boot log of F21 succeeding is here: http://pastie.org/private/ogopev1kbrf5y3lyahhqba

The instances are identical in absolutely every way, except for the image. I have repeated the same result multiple times: F21 always comes up successfully, F22 always fails.

Is this something I should file in a bug report? Is there any way I can get any more debug info on what's happening with F22? (Please keep in mind that because it never brings its networking up, it won't see any user-data scripts I try to inject, nor can I log in to its console.)

—
Erik Swanson

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/users/attachments/20150709/ff1bf7cf/attachment.html>