adding the f command to pf extender view

[wjiec]

In practice, I may be creating port-forward in svc or dp view, but I found that I can't view the enabled forwarding via f command in these views, I have to switch to po view to see it.

So this PR inlines the f command directly in pf_extender, what do you think about this approach?

[derailed]

@wjiec Nice addition! Thank you for this Jayson!!
The question is do we want to surface the pf indicator on these views to clue users that active pfs are in effect (or not) or leave it as is?

[wjiec]

@derailed Yes, that's a good question, I think it would be clearer to add a pf indicator, and it would also help to indicate the current state between different views. Wouldn't it "break" some backward compatibility to add the PF indicator? What do you think?

[derailed]

@wjiec Thanks for your note Jayson!

I think it would be useful to clue users that pfs are active at pod manager level especially in light of these new action.
ie what would prompt you to hit f while in svc or dp in the first place? Seems bubbling this up makes sense.
As for backward compat, the only issue I could foresee is having to change cust view settings to manage this new pf column if users used customization. I don't feel that's a big ask given this affordance.

The stickier point here is to figure out an efficient way to surface status without having to incur an api server call lto find matching pods and hence pfs.
This feels like a heavier lift from this PR original intent.
For this, I think we would need to store pod labels on the pf tracking object and run a match on the viewing resource to assert pf status.

Does this make sense?

[wjiec]

@derailed You are so right! I'm going to try a simple refactoring of the pf tracking object along your lines!

[derailed]

@wjiec Thank you Jayson! We could do a follow up PR for this as what you have here is already a good enhancement...

[wjiec]

@derailed Indeed the subsequent optimization is a rather large modification, so let this PR advanced?

[derailed]

@wjiec Thank you for looking into it! Figured that might be the case ;(