Feat: Rework k8s_tshark functionality and apply the changes in tests

[svteb]

Description

Commit message:

• Prior functionality was bound to fixed time of execution (120s), which introduced problems in testing (tshark session ending before the test began).
• New functionality mainly implements infinite tshark execution along with the possibility of terminating it when deemed appropriate. This is complemented with robust error handling and termination of the tshark process on unexpected crashes during initialization. NOTE: The main tests currently do not handle states where a crash could occur elsewhere and thus a hanging tshark session can still happen (although unlikely).
• The module is properly commented which should allow the user to get a quick understanding of its functionality.
• The user functionality remains the same with easier-to-comprehend function names.
• Handling of PIDs is rather problematic due to the nature of exec_by_node_bg function, which does not return the PID of the tshark process but rather the PID of the shell executing it (unverified). This is why the retrieval of PID may seem rather complicated (especially the pid_command variable). Possible solutions are listed in a comment, but these don't quite work for various reasons (globbing issues, return of incorrect PID, etc.). As for the kill -15 and kill -9 repetition, some tshark session would get stuck in a zombie state if the commands were not executed in this order.

Additional info for reviewers:

The spec tests that verify the correctness of this change are: 5g, oran, core. For some reason I cannot deploy the oran CNF properly so the testing will probably have to be left for github actions. Some of the changes that were made removed the necessity of nesting in the main workload functions (which was thus removed). The class oriented design was somewhat necessary to allow for saving the state of the capture and its future termination. This should also resolve #2072 .

As for the code, a non-trivial amount of time was spent fine-tuning the behavior around PIDs and this is the only configuration I got to behave correctly, although I admit some parts don't look all that pretty. The main problems stem from kubectl_client which is eventually called to spawn the process, since the Process.new/run functions do not return the correct PID/print output I could not really achieve this any other way.

I also discovered that the smf_upf_heartbeat task essentially tests nothing.

Issues:

Refs: #2072 #2087

How has this been tested:

• Covered by existing integration testing
• Added integration testing to cover
• Verified all A/C passes
  • develop
  • master
  • tag/other branch
• Test environment
  • Shared Packet K8s cluster
  • New Packet K8s cluster
  • Kind cluster
• Have not tested

Types of changes:

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update

Checklist:

Documentation

• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• No updates required.

Code Review

• Does the test handle fatal exceptions, ie. rescue block

Issue

• Tasks in issue are checked off

[kosstennbl]

As was discussed online, changes for timeout design and possible nil value of "capture" variable

[kosstennbl]

LGTM
Usage in CNF-Manager is different from usage in tests because of the flawed structure of sample_setup method. For the future - i think we should consider removing this O-RAN E2 connection check from CNF installation, as it seems to be too specific to be in core CNF-Testsuite functionality

[martin-mat]

lgtm

[Smitholi67]

lgtm