Allow multiple nodes have templates (use local storage)

[Atoms]

Issue #, if available:
fixes #451

• Adjusted order of how scheduling happens

  • First we determine if we have usable template (fail fast if not)

• If we have usable template, determine how we need to schedule

  • if Target is provided schedule on that node if template is there
  • if Target and allowedNodes missing assume we want all nodes to use in scheduling (discover all nodes)
  • if templates are on more nodes then is in allowedNodes schedule only on allowedNodes
  • if templates missing on some allowedNodes still continue to schedule on nodes where template exists.

LocalStorage:

• If localstorage (we have no easy way to determine if template is no localstorage):
  • multiple templates can be found
  • assign options.Node same as options.Target as we will use template from same node.
  • templateID and SourceNode cannot be used in this scenario
• If !localstorage
  • only 1 template can be found, if multiple found fail
  • schedule vm on any allowedNode (either Target or allowedNodes or discovered nodes)
  • templateID and SourceNode only can be used in this scenario

Fixed existing tests to reflect multiple scenarios (moving out templateID and SourceNode from setupReconcilerTest

• fixed yamlfmt

createVM had became too complex (gocyclo) as i've written in code for now it should be function to revisit and get into smaller chunks.

[github-actions]

Tests

Please note that running unit and e2e tests requires manual approval from a team member.

e2e tests

We use labels to control which e2e tests contexts are run:

Label	Behaviour
none	Run Generic tests only
e2e/none skip all e2e tests (documentation etc) - overrides all e2e/* labels	Do not run any e2e tests
e2e/flatcar run Flatcar e2e tests	Add Flatcar tests

ℹ️ Ask a team member to add the requested labels if you don't have enough permissions.

[Atoms]

did more intensive testing, not all cases now work, so will be making some changes there.

[Atoms]

fully reworked logic is now pushed, also updated description.

[wikkyk]

Please make better use of validation rules instead of the overriding logic or erroring out. Early failure is better UX.

Make sure the examples are still correct.

This feature deserves documentation :)

This is close. I like the overall idea and implementation but it can be simplified. Please avoid just reformatting unrelated files as that makes the PR larger and thus harder to review plus it can mess up rebasing/reverting/... .

[sonarqubecloud]

Quality Gate passed

Issues
2 New issues
0 Accepted issues

Measures
0 Security Hotspots
82.2% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[Atoms]

Should i rebase on v1alpha2 branch then ?

[wikkyk]

No, that branch is not ready yet. We'll rebase this PR once it's ready. Please hold on tight :)

[mikhailkobik]

Hi guys, any news for this PR? This is most useful feature for homelab usage

[xugxnx]

Hi guys, just to confirm — this feature will come with 0.8.0, right? Any rough ETA for 0.8.0? Really excited!

[wikkyk]

v1alpha2 is close, we should be releasing 0.8.0-alpha.0 soonish, once that is out we will start merging the outstanding PRs like this one.

[wikkyk]

@Atoms v1alpha2 is almost ready. Are you willing to port this? The branch is v1alpha2/wip, it's a bit of a mess but we're not planning more changes to the API.

[65278]

We wanted to include your work in v1alpha2 beta 0, but it didn't make it.
I've rebased your work here: https://github.com/ionos-cloud/cluster-api-provider-proxmox/tree/multiple-node-templates
That said, the rebase is not 100% because I got lost in vm_test.go, amongst other things.

Here's an a priori review:

• You need to use git commit --amend, or git rebase -i main and squash commits. Rebase was harder because of the commit discipline.
• Testing FindVMTemplatesByTags is racy, because golang maps are racy. Redesign the api or the tests.
• I've already removed spec.Target in v1alpha2, but this PR adds it back in, and I didn't 100% the rebase. Please re-remove it (in all commits).

Thanks for your contribution.

[65278]

See my comment.