Create topology controller to manage the lifecycle of Grove ClusterTopology CR and to enable or disable TAS

[sanjaychatterjee]

What you would like to be added?

A valid ClusterTopology CR is a requirement for TAS to be enabled. We need the following:

• Validation checks to ensure ClusterTopology CR is well-formed.
• Validation checks to ensure workload specs align with topology CR
• Define appropriate workflows to enable and disable TAS
• Define appropriate workflows to update the ClusterTopology CR

Why is this needed?

The ClusterTopology CR is used to map the scheduling constraints of a workload to the actual cluster constraints. So, any changes to the CR will impact workloads that are already submitted with topology constraints.

[coderabbitai]

A summary of the changes CodeRabbit can apply:

• Add a ClusterTopology controller and webhook by creating/updating operator/internal/controller/clustertopology/reconciler.go and register.go (reconciler for validation, TAS readiness, periodic revalidation and status updates; register for watches/predicates), operator/internal/webhook/admission/clustertopology/validation/handler.go (create/update/delete validation and immutability enforcement), API types to include ClusterTopologyStatus and condition types, operator config to include ClusterTopologyControllerConfiguration (default concurrentSyncs:1), plus docs (controller README and docs/topology-controller-guide.md) and manifest/codegen updates to enable TAS lifecycle management, validation, and monitoring.

• Implement a production-ready ClusterTopology controller and validation webhook to manage TAS by adding new controller (operator/internal/controller/clustertopology/{reconciler.go,register.go,README.md}), webhook handler and register (operator/internal/webhook/admission/clustertopology/validation/{handler.go,register.go}), docs (docs/topology-controller-guide.md and operator/internal/controller/clustertopology/README.md), and ClusterTopology API status/conditions (operator/api/core/v1alpha1/clustertopology.go), while updating controller configuration types/defaults (operator/api/config/v1alpha1/{types.go,defaults.go}) and wiring registrations (operator/internal/controller/register.go, operator/internal/webhook/register.go), plus adding IMPLEMENTATION_SUMMARY.md.

• ✅ Create PR with these edits
• 📋 Get copyable edits

[kangclzjc]

And maybe we need this controller to provide a mechanism to handle an existing topology like translate the existing topology to Grove ClusterTopology ?