Load Balancer: Add better support for privateLB

[MariusStorhaug]

Change

Minor changes to have the LB be able to be used in a solution deployment. It allows a LB be deployed first, having VMs be deployed after to be put into the backendpool of the LB.

Also added a test for creating a private LB.

Type of Change

Please delete options that are not relevant.

• 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)
• This change requires a documentation update (Wiki)

Checklist

• I'm sure there are no other open Pull Requests for the same update/change
• My corresponding pipelines / checks run clean and green without any errors or warnings
• My code follows the style guidelines of this project
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation (readme)
• I did format my code

[github-actions]

Unit Test Results

    1 files  1 suites   47s ⏱️
    8 tests 8 ✔️   0 💤 0 ❌
104 runs  8 ✔️ 96 💤 0 ❌

Results for commit 85372cf.

♻️ This comment has been updated with latest results.

[MariusStorhaug]

@MrMCake @eriqua ready for a new look. Sorted the comments

[ahmadabdalla]

Should we distinguish the parameter names as:

• 'external.min.parameters.json'
• 'internal.min.parameters.json' (not there yet)
• 'external.parameters.json'
• 'internal.parameters.json'

Similar to the VM module?

[AlexanderSehr]

Should we distinguish the parameter names as:

• 'external.min.parameters.json'
• 'internal.min.parameters.json' (not there yet)
• 'external.parameters.json'
• 'internal.parameters.json'

Similar to the VM module?

@ahmadabdalla would there be a use case for that? In case of the VM the resources & conditions that are deployed differ quite a bit from windows to linux. Is the same true for LB?

[ahmadabdalla]

Not a use case more of one supporting a public ip and one doesn't. If we are introducing a new type, maybe the parameter names should align.

Plus there is a new sku of 'gateway' and tier of 'global' which we are yet to support. This may introduce different properties combinations so we would end up with different files.

[MariusStorhaug]

@ahmadabdalla and @MrMCake: I don't see the need to introduce this now as there is currently no use case for it. We can add it 'when we get there' as its a change that would not affect the module validation or alignment at this point. Just a "lean" thought from my end on this issue. Lets not overthink it at this point.