A bit of an update here:

- We merged the patch in openstack/paunch that stop the Docker container if
we try to start a Podman container.
- We switched the undercloud upgrade job to test upgrades from Docker to
Podman (for now containers are stopped in Docker and then started in
Podman).
- We are now looking how and where to remove the Docker containers once the
upgrade finished. For that work, I started with the Undercloud and patched
tripleoclient to run the post_upgrade_tasks which to me is a good candidate
to run docker rm.

Please look:
- tripleoclient / run post_upgrade_tasks when upgrading
standalone/undercloud: https://review.openstack.org/614349
- THT: prototype on how we would remove the Docker containers:
https://review.openstack.org/611092

Note: for now we assume that Docker is still available on the host after
the upgrade as we are testing things under centos7. I'm aware that this
assumption can change in the future but we'll probably re-iterate.

What I need from the upgrade team is feedback on this workflow, and see if
we can re-use these bits originally tested on Undercloud / Standalone, for
the Overcloud as well.

Thanks for the feedback,


On Fri, Oct 19, 2018 at 8:00 AM Emilien Macchi <emil...@redhat.com> wrote:

> On Fri, Oct 19, 2018 at 4:24 AM Giulio Fidente <gfide...@redhat.com>
> wrote:
>
>> 1) create the podman systemd unit
>> 2) delete the docker container
>>
>
> We finally went with "stop the docker container"
>
> 3) start the podman container
>>
>
> and 4) delete the docker container later in THT upgrade_tasks.
>
> And yes +1 to do the same in ceph-ansible if possible.
> --
> Emilien Macchi
>


-- 
Emilien Macchi
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

Reply via email to