Adopt a Reconcile/Delete pattern for exported methods

[justaugustus]

Describe the solution you'd like
The goals are two-fold here:

1. Adopt a consistent pattern/interface for retrieving/modifying/deleting Cluster/Machine/Azure resources
2. Minimize calls to Azure by storing more information in Cluster & Machine objects

For each actuator reconciler, we currently have Reconcile and Delete methods which operate over the lower-level Azure resources.

Instead of handling implementation details in the top-level reconciler, we should scope them to the resources that they're operating over. As an example, a SubnetSpec should instead be generated/manipulated by a subnet reconciler:

cluster-api-provider-azure/pkg/cloud/azure/actuators/cluster/reconciler.go

Lines 108 to 113 in 0b7cf77

	subnetSpec := &subnets.Spec{
	Name: azure.GenerateControlPlaneSubnetName(s.scope.Cluster.Name),
	CIDR: azure.DefaultControlPlaneSubnetCIDR,
	VnetName: azure.GenerateVnetName(s.scope.Cluster.Name),
	SecurityGroupName: azure.GenerateControlPlaneSecurityGroupName(s.scope.Cluster.Name),
	}

As a concrete reference, see the ReconcileNetwork() method for Cluster API Provider AWS.

A secondary problem is that in our current actuator reconciler model, we make a variety of Create/Update calls to the Azure API, in instances where a Get would suffice. This means we're attempting to modify a majority of the Azure resources we create on every round-trip of a reconciler.

Finally, we do not fully utilize the defined types in the capz API, instead accepting an ambiguous spec and returning a loosely-defined interface, instead of a concrete type.

This means that we are required to query Azure for cluster/machine information, instead of being able to reference the ClusterSpec, ClusterStatus, MachineSpec, and MachineStatus directly, which should be our sources of truth.

---

As a solution, we can:

• Redefine the Service interface to present only Reconcile and Delete methods
• Add a temporary interface (GetterService) to represent packages that don't adhere to the Reconcile/Delete model
• Refactor packages to adhere to Reconcile/Delete model, which will convert and store Azure resource information in the relevant Spec or Status
• Redefine package-level CreateOrUpdate/Gets as unexported methods (only to be used by the package-level Reconcile/Delete)
• Remove references to unexported methods from the top-level reconcilers
• Remove the temporary GetterService interface

The work to accomplish this has begun in #174.

ref: #92
/kind feature

[justaugustus]

/milestone v1alpha1
/priority important-longterm

[justaugustus]

/assign
/remove-help
/priority important-soon
/remove-priority important-longterm

[justaugustus]

/kind design

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[fejta-bot]

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

[k8s-ci-robot]

@fejta-bot: Closing this issue.

In response to this:

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[justaugustus]

/reopen
/remove-lifecycle rotten

[juan-lee]

Have we considered implementing our services as CRDs? @justaugustus

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[rbitia]

/remove-lifecycle rotten

[CecileRobertMichon]

@juan-lee see #416 😃

[CecileRobertMichon]

@justaugustus is it okay if I help out with this?

[CecileRobertMichon]

/assign

[CecileRobertMichon]

/close

in favor of #757

[k8s-ci-robot]

@CecileRobertMichon: Closing this issue.

In response to this:

/close

in favor of #757

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.