[Feature] Redesign workload_resource_test function

[kosstennbl]

For most of the tests we use "workload_resource_test" function to iterate through resources and test required functionality.

Problem is that this function iterates for each container in each resource and sometimes it is not necessary. In some tests, "cnf_workload_resources" function is used for better efficiency, but i don't see that as a good final solution:

• Resource formats provided by these two functions is different, which brings complications and extra adaptation code.
• We don't have a "central point" for resource-related tests, which could bring confusion and instabilities for some changes.

I propose:

• To make "workload_resource_test" an only function for iterations through resources in tests.
• To optimize "workload_resource_test" functionality and usage, let it iterate only required amount of times depending on the test.

[svteb]

It might be better to check what the actual tests are intended to do and what resources they should iterate over instead. Some tests might suffice with access to containers through functions like ClusterTools.all_containers_by_resource?. The library tooling should be revised to give more targeted functions (with sensible names), i.e. get_all_resources, get_all_pods, get_all_containers and they should not be split across two different libraries.

[kosstennbl]

Possible solution would be having option to choose which resources are needed for "workload_resource_test" function, and all the logic behind getting these resources would be hidden from the tests