[systemtest] Upgrade pipeline timeouts because of blocked thread

[im-konge]

Describe the bug

Every run of the upgrade pipelines fails because of the timeout of 360 minutes is hit. This, after my investigation, seems to be issue with the way how we are deleting the topics. Inside our downgrade tests, we are creating 40+ topics. In the end of the test, we are deleting all the resources which are created during the test run. The thing is that we are deleting all those topics one by another - with a few ms gap. Together with deletion, we are doing the waitFor topics to be deleted - so we are using the get() method from Fabric8 K8s client. That's a lot of request, which can end up in some kind of deadlock on the OkHttp side -> because there are too many watchers created etc.

To Reproduce

Steps to reproduce the behavior:

1. Run AZP upgrade pipeline

Expected behavior

The topics should be either deleted as a list or with some kind of poll - 1s. This should fix the issue with deadlock.

Environment (please complete the following information):

• Strimzi version: main
• Installation method: YAML files
• Kubernetes cluster: doesn't matter
• Infrastructure: Minikube