-
Notifications
You must be signed in to change notification settings - Fork 32
feat: Read component descriptor from oci registry #2736
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat: Read component descriptor from oci registry #2736
Conversation
|
❌ Manifests created with 'make dry-run-control-plane' changed! Please make sure to check if changes are needed in related repositories like management-plane-charts, runtime-watcher, etc. |
|
We are blocked here by: |
f5357db to
6fca147
Compare
e258230 to
9d9eb17
Compare
0be33b5 to
3be024d
Compare
071775d to
6aee460
Compare
|
|
d3509af to
5ac7c15
Compare
|
❌ Manifests created with 'make dry-run-control-plane' changed! Please make sure to check if changes are needed in related repositories like management-plane-charts, runtime-watcher, etc. |
1a7339b to
4552c9a
Compare
916d93f to
e3b1e83
Compare
.github/actions/deploy-template-operator-with-modulereleasemeta/action.yml
Show resolved
Hide resolved
76ac242 to
27622a0
Compare
…iptor-from-oci-registry
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I admit that the sheer size of this PR reduces my confidence that everything is okay. It's just hard to really understand everything that is going on. But I think it should be good, we need to get forward at some point.
Let's be careful when this is rolled out. We should explicitly observe how it is working on DEV first. Maybe via a PR-deploy so that it isn't promoted to STAGE automatically.
|
Agree with the previous comment. Let's do an explicit rollout to DEV first via PR deploy. 👍 |
|
And thanks a lot @Tomasz-Smelcerz-SAP for all the effort you put into this! |
f7d93eb
into
kyma-project:main
Description
Note to reviewers: I was really trying to keep this PR as minimal as possible. There are many things to fix/improve/cleanup and many times I've seen that, but I decided not to, in order to introduce as few changes as possible.
Keep that in mind when suggesting improvements. Do we really want to make this PR even bigger? I would rather prefer to make a note in the sources and then create a follow-up PR to clean all the notes. Just remember that the work remaining in #2526 and our other efforts (like simplifying/getting rid of
templatelookup.go) will probably make most of these review-fixes obsolete or unnecessary.Up to you.
Changes proposed in this pull request:
CD) are now read from the OCM repository (which usually is some OCI registry)CD, one must provide an OCM Component Name and a version. In our current setup, the OCM Component Name is available only in aModuleReleaseMetafor a given Module.ModuleReleaseMetabecomes mandatoryCDwas fetched from aModuleTemplateandModuleReleaseMetawas, in some cases, optional. This code requires cleanup in a separate PR. Of course when the test is failing, I had to refactor itModuleReleaseMeta? Currently, when user configures a module on a Kyma, for which there's no ModuleTemplate, the Kyma is put inWarningstate. To keep it consistent, I use the same error->status resolution strategy. To be discussed.templatelookup) decreased. This is because some test cases were disabled. This is (mostly) a legacy code that will disappear anyway. A new method of resolving ModuleTemplates along with ModuleReleaseMeta should be introduced, but this is not a part of this PR. I think it is OK to temporarily "lower the bar" - we're in the middle of a bigger refactoring here. Related issues that will introduce the necessary cleanpup/changes are listed are listed hereRelated issue(s)
#2601