first version of sync-openstack-images
Related to issue automatic OS image activation (capo) (sylva-projects/sylva-core#2013 - closed)
This new Helm chart is introduced for the need of sylva-projects/sylva-core!3565 (merged) (in the context of This MR is introduced in the context of sylva-projects/sylva-core#1101 (closed))
In the context of #1101, we need to have a Helm chart to produce the configmap used as input get-openstack-images tool from sylva-core. This repo here will be the place for this Helm chart, with a new name sync-openstack-images.
As explained in sylva-projects/sylva-core!3565 (merged), this MR only moves the code from sylva-core to this repo, with minor refactoring:
- inputs (Kustomization.spec.postBuild.substitute) are passed via the Values
- the cluster-cloud-config Secret isn't used, the info is now passed via the Values
- the way we handle "Job immutability" changes:
- before this MR the Job was defined by a Flux Kustomization and we had the ability to overcome the immutability of most Job fields by using
Kustomization.spec.force: true - after this MR, nothing equivalent to
force: trueexists in HelmRelease (there is such a field, but there is no ability to delete/recreate), and given that the intention is that set of things which influence the job result will be passed integraly via values, this is replaced by the use of a dynamic job name suffixed by a hash of Helm values
- before this MR the Job was defined by a Flux Kustomization and we had the ability to overcome the immutability of most Job fields by using
Tests: this MR was tested via sylva-projects/sylva-core!3565 (merged)
The work to actually change the behavior to progress in the context of automatic OS image activation (capo) (sylva-projects/sylva-core#2013 - closed) will happen incrementally in follow-up MRs: