Add RVR Status Config Peers Controller

[asergunov]

Add RVR Status Config Peers Controller

Implements #335

Summary

This PR introduces a new controller rvr_status_config_peers_controller that automatically manages peer configuration in ReplicatedVolumeReplica status. The controller ensures that each replica has up-to-date information about all other replicas (peers) in the same ReplicatedVolume, excluding itself.

Changes

New Controller Implementation

• Controller: rvr_status_config_peers_controller
  • Watches ReplicatedVolume resources
  • Watches ReplicatedVolumeReplica resources (via owner reference)
  • Updates status.config.peers field in each ReplicatedVolumeReplica

Key Features

1. Peer Discovery: Automatically discovers all replicas belonging to a ReplicatedVolume by:

   • Filtering replicas controlled by the ReplicatedVolume
   • Validating that replicas have required fields (NodeName, Status.Config.NodeId, Status.Config.Address)
   • Building a peer map keyed by node name

2. Peer Management:

   • Each replica's status.config.peers contains all other replicas (excluding itself)
   • Handles edge cases:
     • Single replica scenarios (no peers)
     • Multiple replicas on the same node (only first one is kept as peer)
     • Replica deletion (automatically removes deleted replicas from peers)
   • Includes Diskless flag in peer information

3. Optimization:

   • Only patches replicas when peer configuration actually changes
   • Uses status subresource patching for efficient updates

Testing

Comprehensive test suite using Ginkgo/Gomega covering:

• Basic peer configuration scenarios
• Edge cases (single replica, missing fields, invalid owner references)
• Error handling (Get, List, Patch failures)
• Multiple replicas scenarios
• Replica deletion
• Diskless replica handling
• Idempotency (no unnecessary updates)

Dependencies

• Updated github.com/onsi/ginkgo/v2 to v2.27.2
• Updated github.com/onsi/gomega to v1.38.2
• Updated various golang.org/x packages

Files Changed

• images/controller/internal/controllers/rvr_status_config_peers/controller.go - Controller setup
• images/controller/internal/controllers/rvr_status_config_peers/reconciler.go - Reconciliation logic
• images/controller/internal/controllers/rvr_status_config_peers/reconciler_test.go - Test suite
• images/controller/internal/controllers/rvr_status_config_peers/rvr_status_config_peers_suite_test.go - Test suite setup
• images/controller/internal/controllers/registry.go - Controller registration
• images/controller/go.mod - Dependency updates
• images/controller/go.sum - Dependency checksums

How It Works

1. When a ReplicatedVolume or its ReplicatedVolumeReplica changes, the controller is triggered
2. The controller:
   • Fetches the ReplicatedVolume
   • Lists all ReplicatedVolumeReplica resources
   • Filters to only include replicas controlled by the ReplicatedVolume with required status fields
   • Builds a peer map from the filtered replicas
   • For each replica, updates its status.config.peers with all other replicas (excluding itself)
   • Only patches replicas when the peer configuration has changed

Benefits

• Automatic Peer Discovery: Replicas automatically know about each other without manual configuration
• Consistency: Ensures all replicas have consistent peer information
• Resilience: Automatically handles replica additions and deletions
• Efficiency: Only updates when necessary, reducing unnecessary API calls