Description of problem: The downstream only strange patch applied for fixing https://bugzilla.redhat.com/show_bug.cgi?id=1466744 made very difficult and fragile to do non-containerized setups. Please provide a `stable` easy way to create old fashioned setups for example by adding on option (env var?) do skip the environment file adding done by triple-common, or provide an environment file in the THT which negates the effect of docker.yml Version-Release number of selected component (if applicable): openstack-tripleo-heat-templates-7.0.0-0.20170913050524.0rc2.el7ost.noarch openstack-puppet-modules-11.0.0-0.20170828113154.el7ost.noarch instack-undercloud-7.4.1-0.20170912115418.el7ost.noarch How reproducible: always Actual results: Deployment fails becuse the images to use are not specified (it is normally an extra env file, which passed in case of docker deployments) Additional info: I am ok with reverting the changes made for rhbz#1466744 .
If non-containerized deployments are *non-supported* by redhat in osp12, why would we want to invest time/resources in testing unsupported environments ?
Created attachment 1333778 [details] docker-to-baremetal.sh script Script to convert the tripleo-heat-template defaults back to baremetal.
See the attached script above which can be used to flip the OSP12 docker defaults back to baremetal. You would need to execute this script on a set of tripleo-heat-templates used for deployment before running the deployment commands.
So I think Bug 1496782 is also related to this as you should be able to do a deployment with the mixed versions which is blocked by this. Ocata won't have the docker yaml file which causes it to fail. We'll need a way to skip this for mix-version deployments
Alex: scope of this bug was OSP12 Pike only I think.
Right but OSP12 undercloud should be able to deploy OSP11 templates. See Bug 1496782 for context
Alex: To clarify, knowing the implementation details here I would like to treat these as two separate bugs for now I think. The patch which satisfies this bug probably won't help with the other bug you've linked in.
That's fine. I was just pointing out that the same thing is perhaps causing both issues and rather than treating them as separate, perhaps it makes sense to combine the fix(es) since the root cause is most likely the same for both. I was mostly just commenting for awareness.
Comment on attachment 1333778 [details] docker-to-baremetal.sh script This version of the script got munged. New version will upload now.
Created attachment 1333944 [details] docker-to-baremetal.sh script
Sasha: thanks for the feedback on the bad docker-to-baremetal.sh script I initially uploaded. The new one should be good.
Worked.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2017:3462