feat: Add node-not-ready event monitoring support

[kyungho-for-ops]

Overview

This PR adds support for monitoring Kubernetes Node node-not-ready events in khook, enabling the system to detect and respond to node-level issues automatically.

What Changed

✨ New Features

• Added node-not-ready event type to supported event configurations
• Implemented dedicated node event processing logic (mapNodeEventType)
• Enhanced event filtering to handle Node and Pod events separately

🔧 Technical Changes

• API Changes: Extended EventConfiguration enum with node-not-ready type
• Event Processing: Added mapNodeEventType function for node-specific events
• Event Filtering: Modified event type mapping to distinguish Pod vs Node events
• CRD Updates: Updated Kubernetes CRD schemas and Helm chart CRDs
• Code Generation: Updated auto-generated deepcopy code

📚 Documentation Updates

• Updated README.md with node-not-ready event type documentation
• Added comprehensive node recovery agent example configuration
• Enhanced event types table with node monitoring capabilities

Files Modified

api/v1alpha2/hook_types.go              # Added node-not-ready validation
internal/event/watcher.go               # Implemented node event detection  
README.md                               # Updated docs and examples
config/crd/bases/kagent.dev_hooks.yaml  # Updated CRD schema
helm/khook-crds/crds/kagent.dev_hooks.yaml # Updated Helm CRD
api/v1alpha2/zz_generated.deepcopy.go  # Auto-generated updates
.gitignore                              # Excluded build scripts

Usage Example

apiVersion: kagent.dev/v1alpha2
kind: Hook
metadata:
  name: node-monitoring-hook
spec:
  eventConfigurations:
  - eventType: node-not-ready
    agentId: node-recovery-specialist
    prompt: |
      CRITICAL: Node {{.ResourceName}} is not ready at {{.EventTime}}.
      
      AUTONOMOUS MODE: Diagnose and resolve node issues immediately:
      • Check node conditions (Ready, MemoryPressure, DiskPressure, PIDPressure)
      • Analyze kubelet logs and system resources
      • Verify network connectivity and DNS resolution
      • Attempt node recovery procedures
      • If recovery fails, safely drain and replace node
      • Annotate fixed resources with: kagentFix=<timestamp>

Testing

• Existing Pod event processing functionality remains intact
• Node event detection logic implemented and validated
• CRD validation rules updated and tested
• Backward compatibility maintained

Breaking Changes

❌ None - This is a purely additive feature maintaining full backward compatibility.

Type of Change

• New feature (non-breaking change which adds functionality)
• Documentation update

[antweiss]

Looks fine to me. Just need to remove the redundant .sh file in helm requirements.lock and sign off the commit

[antweiss]

Why is this needed?

[kyungho-for-ops]

it dosen't necessary.
i'll delete it

[antweiss]

Thank you for this! Also - if you have any thoughts on how to make new event support more streamlined - LMK.

[antweiss]

hi @kyungho-for-ops - waiting for you to sign this off.

[kyungho-for-ops]

Hi @antweiss

I've rebased the branch using git rebase HEAD~4 --signoff and force-pushed to fix all DCO issues. All commits should now be signed off correctly.

Could you please approve the pending workflow so the CI/CD checks (Docker Build, Test) can start? Once the checks pass, I would appreciate your final approving review for merging. Thank you for your patience!

[antweiss]

The last commit (trigger final checks) isn't signed-off :))

[kyungho-for-ops]

Sorry for the late response; I just got back last night. I've finished the changes.