Skip to content

Update currently built containers - #111

Open
elfiesmelfie wants to merge 1 commit into
openstack-k8s-operators:mainfrom
elfiesmelfie:efoley_update_readme
Open

Update currently built containers#111
elfiesmelfie wants to merge 1 commit into
openstack-k8s-operators:mainfrom
elfiesmelfie:efoley_update_readme

Conversation

@elfiesmelfie

Copy link
Copy Markdown
Contributor

README and containers/image-mappings are out of date

The following commands were used to create the new lists:

README:
find -name Containerfile | xargs dirname {} | awk -F/ '{print "- openstack-"$NF";"}' | sort

containers/image-mappings:
find -name Containerfile | xargs dirname {} | sed 's:^./containers/::' | sort | awk '{print " "$NF":"}'

get a list of openstackversions image vars
oc explain openstackversions.spec.customContainerImages | grep Image | awk '{ print $q}'

@openshift-ci

openshift-ci Bot commented Aug 19, 2026

Copy link
Copy Markdown

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign amoralej for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details Needs approval from an approver in each of these files:

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

neutron/neutron-metadata-agent-ovn:
- edpmNeutronMetadataAgentImage
neutron/neutron-ovn-agent:
- edpmNeutronOvnAgentImage

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have omitted neutron/neutron-server because I couldn't find the var in openstackversions.spec.customContainerimages.
Please someone let me know the right value!

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

it should be

    neutron/neutron-server:
      - neutronAPIImage

but a confirmation from @openstack-k8s-operators/network team is totally appreciated :)

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes correct neutronAPIImage to be used, but we can't add it now, can add it as commented and once neutron-operator is adapted to support these eventletless images we can uncomment. Planning to work on that next week.

@amoralej amoralej Aug 20, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@karelyatin That means we can't deploy operators in current state with these s2i images? If so, you plan in doing the changes in the operator backwards compatible?

BTW, adding it to this file has actually not impact, I think, other than having it ready for future CI jobs.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

<< @karelyatin That means we can't deploy operators in current state with these s2i images? If so, you plan in doing the changes in the operator backwards compatible?
@amoralej plan is to make it such that greenfield deploy run with master/19 images, while handling the minor update case where deployment strategy change only during minor update but not when operator rollouts happen. This may require some CI hacks until fully migrated so will take care of those.
Or you mean we need to also support running antelope images during greenfield install as well?

<< BTW, adding it to this file has actually not impact, I think, other than having it ready for future CI jobs.
From recent meetings i understood CI jobs preparation is in progress and AFAIU this file will be used to detect available images, and if this is true we can't add neutron-server yet there until we adapt neutron-operator.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This may require some CI hacks until fully migrated so will take care of those. Or you mean we need to also support running antelope images during greenfield install as well?

@karelyatin My concern was the transition and avoid dependency between changing neutron-operator and updating defaults and CI to use new images. Making the change backwards compatible may help for transition I guess but if there is any way to avoid those dependencies that may block moving to new containers or may add dependency to other projects switch, no problem.

WRT update from 18 to 19. I'm not very familiar with the RHOSO update process but after initial update of the operators only, wouldn't the operators run with the former openstackversion as an intermediate step? Is the update of operator and openstackversion atomic? I'm not sure how that's is handled tbh.

i understood CI jobs preparation is in progress and AFAIU this file will be used to detect available images

All that is right, it was just about timing. I though we are not in the status to run that yet (actually, iiuc deployment will fail on neutron-server anyway) although I may be wrong.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@karelyatin My concern was the transition and avoid dependency between changing neutron-operator and updating defaults and CI to use new images. Making the change backwards compatible may help for transition I guess but if there is any way to avoid those dependencies that may block moving to new containers or may add dependency to other projects switch, no problem.

It would be kind of backward compatible only for the upgrade scenario i.e neutron pod will keep running with operator updates to support new way. The only thing is greenfield/fresh setup to use new images always and switch to new method during update. For CI we may need some hacks to switch default images as they are managed across different repos, so will see that once i push that next week

WRT update from 18 to 19. I'm not very familiar with the RHOSO update process but after initial update of the operators only, wouldn't the operators run with the former openstackversion as an intermediate step? Is the update of operator and openstackversion atomic? I'm not sure how that's is handled tbh.

During update:-

  • First stage operators are updated, which provides newer OpenStackVersion and all service operators are updated, this can result in rolling out changes to already deployed services . So in this stage, neutron and other pods will keep running/functional with older container images, there may be some restarts though
  • To trigger minor update(in this case it could be major version though but not have any visibility on plan for these in general), spec.targetVersion is updated, at this pods will be rolled out with new(s2i) images and at this stage we will switch to the the new deploy strategy for neutron. Add ServiceDefaults mechanism for updates support openstack-operator#1439 should provide some picture on how it's being done, it's already used for glance/rabbitmq/manila transitions

Comment thread containers/image-mappings.yaml
@amoralej

Copy link
Copy Markdown
Contributor

I was thinking in updating it at the end to avoid forcing ongoing PRs to rebase under the assumption that this is not used yet but we can merge it if needed.

@elfiesmelfie

Copy link
Copy Markdown
Contributor Author

I was thinking in updating it at the end to avoid forcing ongoing PRs to rebase under the assumption that this is not used yet but we can merge it if needed.

I added the README and containers/image-mapping updates to the developer docs, so hopefully, we have this included in PRs from now on and this won't be an ongoing update.

@centosinfra-prod-github-app

Copy link
Copy Markdown

Build failed (check pipeline). Post recheck (without leading slash)
to rerun all jobs. Make sure the failure cause has been resolved before
you rerun jobs.

https://gateway-cloud-softwarefactory.apps.ocp.cloud.ci.centos.org/zuul/t/rdoproject.org/buildset/4260294e387d4ebe99931a5b5d14332d

✔️ s2i-openstack-containers-molecule SUCCESS in 2m 08s
s2i-openstack-container-content-provider FAILURE in 4m 26s

README and containers/image-mappings are out of date

The following commands were used to create the new lists:

README:
find -name Containerfile | xargs dirname {} | awk -F/ '{print "- `openstack-"$NF"`;"}' | sort

containers/image-mappings:
find -name Containerfile | xargs dirname {} | sed 's:^./containers/::' | sort | awk '{print "    "$NF":"}'

get a list of openstackversions image vars
oc explain openstackversions.spec.customContainerImages | grep Image | awk '{ print $q}'
@openshift-ci

openshift-ci Bot commented Aug 24, 2026

Copy link
Copy Markdown

PR needs rebase.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants