Open Stack

On Wed, Jun 29, 2016, at 10:41 AM, Clint Byrum wrote:
> Excerpts from Clark Boylan's message of 2016-06-28 13:23:25 -0700:
> > It is an even year again which means there is a new Ubuntu LTS release
> > out (Xenial). We currently have Xenial  images building and uploaded and
> > the Openstack Ansible and Openstack Puppet groups are both taking
> > advantage of them. Our "base" set of devstack jobs also seem happy on
> > Xenial as well (see the experimental results on
> > https://review.openstack.org/#/c/330835/1). All this to say I think we
> > are ready to do the big switch and get everyone running on Xenial by
> > default.
> > 
> 
> I don't know if this has been done in the past, but it seems like there
> should be a bit of a lead-time for openstack-dev and openstack-operators
> to know they need to start working on Xenial. Doing it in the middle of
> a dev cycle seems like it would explode the install guide at a pretty
> critical moment, vs. just making sure everyone knows it will be done on
> day 1 of Ocata, and then allowing summit plans to be made for doc
> updates,
> etc. etc.

I disagree for several reasons:

* We have done this many times in the past while this alone is not a
good reason it hasn't been an issue for operators or developers
previously. We do try to put a good deal of effort into the transition
to make sure it goes as smoothly as possible.
* Devs expect this switch to happen soon. In part due to the previous
point but also because many of them want newer versions of software.
Kolla needs new kernel for filesystem things, Neutron always wants newer
OVS, Nova wants newer libvirt, and so on.
* On the ops front both the Ansible and Puppet groups have already taken
the plunge. The operators are ahead of development and we should catch
up.
* I am not sure how this is related to the install guide. Andreas
mentioned we could possibly run all the doc builds on Xenial including
those for existing stable branches. As long as the current guide builds
what is the concern?
* This reason is a bit selfish but I would really like to avoid
extending Trusty's lifetime in our systems another 6 months.
Particularly since Trusty is upstart and Xenial is systemd. The quicker
we make that transition and move on the better.
* (TC Correct me if I am wrong) TC has stated we should support Current
Ubuntu LTS. This maneuver gets us inline with that statement.

We really do try to make this happen as smoothly as possible. I won't
say it will be perfect because history says we will likely uncover a
corner case or two (broken python3.4 on trusty anyone), but it shouldn't
cause major issues. Even if it does reverting is straightforward.

Clark