Add functionality and performance tests based on kci-dev

[tales-aparecida]

https://github.com/kernelci/kci-dev is the official python client for the dashboard, and it should work as well as the web-UI.

Let's create some GitHub pipeline evaluating it is working as best as it can

[MarceloRobert]

@aliceinwire I see that the github CI for kci-dev has a couple of tests, what kind of tests would you like to bring to dashboard? We can work on handshaking dashboard and kci-dev.

A thing that I think is also useful is what I talked about in the Weekly 2 weeks ago about kci-dev downloading the dashboard api schema and checking for changes

[tales-aparecida]

aliceinwire I see that the github CI for kci-dev has a couple of tests, what kind of tests would you like to bring to dashboard? We can work on handshaking dashboard and kci-dev.

I think we can pretty much run the same test suite pre-merge. Not necessarily gating on it, but at least an informative run.

[tales-aparecida]

In GitLab I know we can trigger multi-project pipelines, which help a lot with API/integration testing, which is basically "trigger the pipeline from this other repo". I'm not familiar how that works in GitHub

[tales-aparecida]

I don't think we need to work on such refinement soon. The easiest path, as a proof of concept, would be to simply copy kci-dev test suite to the dashboard repo, get some fixtures and run the test suite "offline". If we want, we can work on removing duplication with fancy GitHub triggers in the future

[tales-aparecida]

The reported latest example of a breaking change: kernelci/kci-dev@2d6022e

[tales-aparecida]

Before the issue above, part of the pain comes from using staging:9000 to in the kci-dev CI. It was failing quite frequently due to the issues with hw-listing performance, and that brought:

A thing that I think is also useful is what I talked about in the Weekly 2 weeks ago about kci-dev downloading the dashboard api schema and checking for changes

Which we settled as a great idea for the future of kci-dev CI, along with running it offline (mocking the API instead of effectively running integration testing every time) and to verify the latest schema, to prepare against minor changes.