Re: [atomic] OpenStack Magnum concerns about Project Atomic

[Nick Coghlan <ncoghlan gmail com>]

On 26 February 2016 at 02:17, Matthew Miller <mattdm mattdm org> wrote:
> On Wed, Feb 24, 2016 at 11:40:10AM -0500, Matt Micene wrote:
>> CoreOS is a standalone distro, and can move at whatever speed for whatever
>> component they like.  Our packages move at the speed of Fedora.  The Atomic
>> members may be maintainers of a component, but if Fedora decides that, for
>> example, docker 1.10 is a breaking change that needs to wait until F24,
>> there's not a lot we can do about it for a F23 based Atomic Host.
>
> This is basically true, but we're also trying to figure out how to do
> it better. There are several options:
>
> * Simply packaging these things for parallel existence, like docker110
>   (not my preference, but easily done with no changes to the way we do
>   anything)
>
> * Putting packages like this in updates-testing, but *not* pushing them
>   to stable; pulling from updates-testing for the Atomic tree compose

1.10 is there in F23 updates-testing already - the problem with
pushing it to stable is the image migration which *did* break the
docker service on my machine.

Upgrading my particular system required:

    sudo dnf install --enablerepo=updates-testing docker docker-v1.10-migrator
    sudo v1.10-migrator-local -s devicemapper # Not a typo, there's no
docker prefix
    sudo systemctl start docker

So in this particular case, even if you did allow the compose to draw
from updates-testing, I'm guessing Atomic Host would face the same
problem I did on Workstation: how can you automatically handle the
format change for the image store for previously deployed hosts?

(This will also be a problem to be aware of for F23 -> F24 system upgrades)

Cheers,
Nick.

-- 
Nick Coghlan   |   ncoghlan gmail com   |   Brisbane, Australia