[openstack-dev] [neutron] multinode CI jobs in the gate

John Schwarz jschwarz at redhat.com
Sun Dec 18 11:28:23 UTC 2016


There is already an experimental job called
(which was added by [1]).

However, the job is currently a "one dvr_snat nodes, two dvr nodes"
deployment (instead of the "two dvr_snat nodes, one dvr node" that it
should be in order to make L3HA work there). The patch which makes the
final transition is being worked on in [2], and it's ready to be
merged afaik. Once it merges, we'll have a DVR+HA gate in the
experimental queue.


[1]: https://review.openstack.org/#/c/383742/
[2]: https://review.openstack.org/#/c/383827/

On Fri, Dec 16, 2016 at 2:23 AM, Armando M. <armamig at gmail.com> wrote:
> Hi Neutrinos,
> Infra patch proposed in [1] got me thinking again about what we shall do
> when it comes to multinode testing in the gate and how to focus our testing
> and CI efforts upstream going forward. My line of thinking has always been
> that multinode resources should be devoted to configurations whose coverage
> fully benefit from the inherent nature of the distributed deployment, and
> that means giving priority to DVR+HA and demote other configurations that
> use centralized routing.
> I know that some of you in team have worked on this, but I am a bit behind
> on the latest status update. Any good soul willing to bring a strapped (for
> time) PTL up to speed?
> Cheers,
> Armando
> [1] https://review.openstack.org/#/c/411263/
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

John Schwarz,
Senior Software Engineer,
Red Hat.

More information about the OpenStack-dev mailing list