[e2e] Test have been failing for a long time

[richardcase]

What happened:
Our nightly e2e test have been failing for a long time now.

What did you expect to happen:
The tests to pass consistently

How to reproduce it:

Anything else you would like to add:
[Miscellaneous information that will assist in solving the issue.]

Environment:

• rke provider version:
• OS (e.g. from /etc/os-release):

[Danil-Grigorev]

@furkatgofurov7 To fix this we need to fix #263. I propose to do this in a separate PR and merge #260 as is, as it resolves contract issue.

[Danil-Grigorev]

Also, kubernetes-sigs/cluster-api#9952 seems to be required here too.

[Danil-Grigorev]

@richardcase I think this was closed by mistake. e2e tests are not passing with this merged change alone for me.

[richardcase]

They passed for me this morning

[Danil-Grigorev]

@furkatgofurov7 You mentioned there is an issue with the e2e tests, later in the run. Can you give more details? I’ll retry on my side as well. In my setup I was hitting originally this issue: #263

[richardcase]

Maybe this is now hinting an environmental factors. I will run it in different environments to check.

I also created #267 so we can run the e2e via self hosted runners, to test.

[furkatgofurov7]

@furkatgofurov7 You mentioned there is an issue with the e2e tests, later in the run. Can you give more details? I’ll retry on my side as well. In my setup I was hitting originally this issue: #263

It was running in my fork, re-running it again to test the latest changes https://github.com/furkatgofurov7/cluster-api-provider-rke2/actions/runs/7958544104/job/21723628664

[Danil-Grigorev]

I see the job is passing green. It looks like it is environment specific in some way: https://github.com/rancher-sandbox/cluster-api-provider-rke2/actions/runs/7958280349

[furkatgofurov7]

@furkatgofurov7 You mentioned there is an issue with the e2e tests, later in the run. Can you give more details? I’ll retry on my side as well. In my setup I was hitting originally this issue: #263

It was running in my fork, re-running it again to test the latest changes https://github.com/furkatgofurov7/cluster-api-provider-rke2/actions/runs/7958544104/job/21723628664

Passed in my fork as well: https://github.com/furkatgofurov7/cluster-api-provider-rke2/actions/runs/7958544104

[Danil-Grigorev]

Yes, it passed for me as well. There was a change I overlooked - changed number of nodes from 1 to 3. This corresponds with a workaround for the bug intended to be fixed #263 and thus e2e does not hit the problem anymore.