User or Developer story

As a developer I want to integrate the Keda upstream e2e tests in the custom-metrics-autoscaler-operator repository so that problems can be detected before the changes go to the internal pipeline.

Description

zroubali@redhat.com feel free to add more color to the description of the ticket

Engineering Details

https://github.com/kedacore/keda/tree/main/tests
we should run: internals, and from the scalers package also: prometheus, kafka, cpu, memory
this is how you run the tests: they expect kubernetes cluster, then the setup installs upstream KEDA (we should modify this to use our CMA build) and then tests are running against this instance https://github.com/kedacore/keda/tree/main/tests
we can have some variation of this runner modified for our purposes: https://github.com/kedacore/keda/blob/main/tests/run-all.go

• sequential are actually nice to have as well, but they actually modifies KEDA installation, so shouldn’t be run in parallel with other tests

Acceptance Criteria

The following e2e tests run on every pull request and are required

• `internals`
• from the `scalers` package also: prometheus, kafka, cpu, memory
• `sequentia` test actually modifies KEDA installation, so they should run in a different e2e test suit