Open Stack

Hello,

In our weekly IRC meeting, all developers and others in attendance were in unanimous agreement that upgrades are the #1 priority for Kolla to solve [1].  inc0 has developed both a simple database migration implementation  around keystone [2] as well as a more complex playbook around nova [3].  Keystone has merged - nova looks really solid but needs a little more attention.

We have 10 core reviewers on the Kolla core review team.  I would ask that each core reviewer choose 1 or 2 services to solve upgrades on and spend 40 hours out of the next two weeks prior to our midcycle implementing the work items they have committed themselves to.  If you are not a core reviewer but feel you want to get involved in this work, please assign yourself to a maximum of one service and commit as described below.

The work for OpenStack services should be copy/paste/search/replace/test from inc0's work.  The work for infrastructure services (haproxy keepalived etc) is more R&D based around defining what needs to be done for these services.  If you take two playbooks, I recommend taking one of each type of service (infrastructure and an Openstack service).  Neutron, openvswitch, cinder and ceph will be the most difficult services remaining to upgrade, so if your in for a good challenge, pick up one of those ;)

To register for the 40 hour time commitment related to this critical development effort do the following:

  *    ping me on IRC in the kolla IRC channel with a statement indicating you are able to commit 40 hours out of the next 2 weeks to focus on this critical development effort
  *   Find the appropriate upgrade-servicename and set yourself as the assignee
     *   https://launchpad.net/kolla/+milestone/mitaka-3
  *   identify yourself with a service under work items here:
     *   https://blueprints.launchpad.net/kolla/+spec/upgrade-kolla

The reason for the 40 hours?  I recognize I'm asking for a big time commitment from our community, but this effort is critical, critical, critical.  I want the attendees at the midcycle fully ramped up on our upgrades implementation.  If we don't deliver upgrades in Mitaka, it will delay deployment of Kolla because there wont be a migration strategy between different versions.  That is why [1] was unanimous on upgrades ;)  This effort is critical for project progression.

Regards,
-steve

[1] http://eavesdrop.openstack.org/meetings/kolla/2016/kolla.2016-01-13-16.30.log.html#l-245
[2] https://blueprints.launchpad.net/kolla/+spec/upgrade-keystone
[3] https://blueprints.launchpad.net/kolla/+spec/upgrade-nova
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20160123/e8384e43/attachment.html>