On Mon, Nov 28, 2016 at 09:09:09AM +0200, Barak Korren wrote:
On 28 November 2016 at 09:03, Dan Kenigsberg
<danken(a)redhat.com> wrote:
>
> Adding Petr; I assume he won't have plenty of time to fix the issue, we
> should mark it as broken. Any test with a timeout is broken-on-ci by
> design.
>
What exactly is timing out? Can we help to fix that? Is it waiting
forever for something that will never "show up" because it is outside
the mock env (And could be bind-mounted into it)?
I did not research this deeply, but in that test we create network
interfaces, and wait for the ensuing netlink events. If they don't
arrive within 2 seconds, we declare failure.
The failure we see is what I'd expect from a very slow slave.