feat: add initial setup for portal and cx-iam

[evegufy]

Description

This PR adds the helm charts from portal and cx-iam (versions from 3.2/23.09 release) to the umbrella chart.
It doesn't contain tests in this sense and for a proper proper e2e testing functional tests as wells as more config would need to get added, still, for an initial integration in the umbrella helm chart this PR should be sufficient.

Pre-review checks

Please ensure to do as many of the following checks as possible, before asking for committer review:

• DEPENDENCIES are up-to-date. Dash license tool. Committers can open IP issues for restricted libs.
• Copyright and license header are present on all affected files

[carslen]

PR looks straight forward. Are these failing tests ok, @almadigabor ? I don't get the reason for the failure from error messages for both failed checks.

[almadigabor]

I don't think the errors are related to the pull request. Kyverno fails as some of the dependencies have not set runAsNonRoot in the pod specification. Helm test fails with the edc dataplane pod if I get it correctly.

[evegufy]

@carslen @almadigabor thanks for reviewing! Last weeks I successfully tested the install on my local Minikube.
IMO the issue with ct here, maybe upgrading the helm testing action would be worth a try or replacing it entirely with a simple helm install.

[fty4]

I just tried to start this branch locally in my minikube and had following problems:

• two pods will keep restarting (processes-worker and e2e-sharedidp)
• some pods does not have the given Helm install name as prefix

IMO the issue with ct here, maybe upgrading the helm testing action would be worth a try or replacing it entirely with a simple helm install.

Regarting to your last comment, @evegufy: I do not think we should replace the ct with a helm install - this test framework is quiet common and will print a detailed log.

For example: in the last run of this pr you can see that the runner might be to small to run that heavy load:0/1 nodes are available: 1 Insufficient cpu. preemption: 0/1 nodes are available: 1 No preemption victims found for incoming pod.

We might need to use larger runner or try to limit the load.

[evegufy]

I just tried to start this branch locally in my minikube and had following problems:

• two pods will keep restarting (processes-worker and e2e-sharedidp)
• some pods does not have the given Helm install name as prefix

IMO the issue with ct here, maybe upgrading the helm testing action would be worth a try or replacing it entirely with a simple helm install.

Regarting to your last comment, @evegufy: I do not think we should replace the ct with a helm install - this test framework is quiet common and will print a detailed log.

For example: in the last run of this pr you can see that the runner might be to small to run that heavy load:0/1 nodes are available: 1 Insufficient cpu. preemption: 0/1 nodes are available: 1 No preemption victims found for incoming pod.

We might need to use larger runner or try to limit the load.

@fty4 Thanks for testing!

Regarding 'processes-worker': that's cronJob, and it's normal that it restarts every couple of minutes
Regarding 'sharedidp': what's your local OS? I tested on Ubuntu. If you're using Mac with Apple Silicon, the reason for the restarts could be the mismatching cpu arch. The image is only available for amd64, in this case it can help to provide more cpu cores to your Minikube to enable proper emulation of the different cpu arch, but the performance on Mac will even then probably not be great and the pod will need a couple of restarts to start up.

Good hint regarding the insufficient cpu. We probably need to add a config for the KinD action as the default values aren't sufficient.

I'm still not quite convinced if ct is really suitable for testing such a chart:
When having enabled both keycloak instances (sharedidp and centralidp), I encountered for instance this error which I wasn't encountering on my local Minikube.
Also, in another context, I already ran into helm/chart-testing#332 but this could be not an issue if we upgrade helm testing action.
Still, I'm open to try some other things instead of removing ct.