Hi Spyros, Thanks for posting the discussion output. I'm not sure I can follow the idea of simplifying CNI configuration. Though we have both calico and flannel for k8s, but if we put both of them into single one config script. The script could be very complex. That's why I think we should define some naming and logging rules/policies for those scripts for long term maintenance to make our life easier. Thoughts?
On 25/06/18 19:20, Spyros Trigazis wrote: > Hello again, > > After Thursday's meeting I want to summarize what we discussed and add > some pointers. > > * Work on using the out-of-tree cloud provider and move to the new > model of defining it > https://storyboard.openstack.org/#!/story/1762743 > <https://storyboard.openstack.org/#%21/story/1762743> > https://review.openstack.org/#/c/577477/ > * Configure kubelet and kube-proxy on master nodes > This story of the master node label can be > extened https://storyboard.openstack.org/#!/story/2002618 > <https://storyboard.openstack.org/#%21/story/2002618> > or we can add a new one > * Simplify CNI configuration, we have calico and flannel. Ideally we > should a single config script for each > one. We could move flannel to the kubernetes hosted version that > uses kubernetes objects for storage. > (it is the recommended way by flannel and how it is done with kubeadm) > * magum support in gophercloud > https://github.com/gophercloud/gophercloud/issues/1003 > * *needs discussion *update version of heat templates (pike or > queens) This need its own tread > * Post deployment scripts for clusters, I have this since some time > for my but doing it in > heat is slightly (not a lot) complicated. Most magnum users favor > the simpler solution > of passing a url of a manifest or script to the cluster (at least > let's add sha512sum). > * Simplify addition of custom labels/parameters. To avoid patcing > magnum, it would be > more ops friendly to have a generic field of custom parameters > > Not discussed in the last meeting but we should in the next ones: > > * Allow cluster scaling from different users in the same project > https://storyboard.openstack.org/#!/story/2002648 > <https://storyboard.openstack.org/#%21/story/2002648> > * Add the option to remove node from a resource group for swarm > clusters like > in kubernetes > https://storyboard.openstack.org/#!/story/2002677 > <https://storyboard.openstack.org/#%21/story/2002677> > > Let's follow these up in the coming meetings, Tuesday 1000UTC and > Thursday 1700UTC. > > You can always consult this page [1] for future meetings. > > Cheers, > Spyros > > [1] https://wiki.openstack.org/wiki/Meetings/Containers > > On Wed, 20 Jun 2018 at 18:05, Spyros Trigazis <[email protected] > <mailto:[email protected]>> wrote: > > Hello list, > > We are going to have a second weekly meeting for magnum for 3 weeks > as a test to reach out to contributors in the Americas. > > You can join us tomorrow (or today for some?) at 1700UTC in > #openstack-containers . > > Cheers, > Spyros > > > > > __________________________________________________________________________ > OpenStack Development Mailing List (not for usage questions) > Unsubscribe: [email protected]?subject:unsubscribe > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev -- Cheers & Best regards, Feilong Wang (王飞龙) -------------------------------------------------------------------------- Senior Cloud Software Engineer Tel: +64-48032246 Email: [email protected] Catalyst IT Limited Level 6, Catalyst House, 150 Willis Street, Wellington --------------------------------------------------------------------------
__________________________________________________________________________ OpenStack Development Mailing List (not for usage questions) Unsubscribe: [email protected]?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
