Vapora/scripts/deploy-k8s.nu

131 lines
4.4 KiB
Plaintext
Raw Permalink Normal View History

feat: Phase 5.3 - Multi-Agent Learning Infrastructure Implement intelligent agent learning from Knowledge Graph execution history with per-task-type expertise tracking, recency bias, and learning curves. ## Phase 5.3 Implementation ### Learning Infrastructure (✅ Complete) - LearningProfileService with per-task-type expertise metrics - TaskTypeExpertise model tracking success_rate, confidence, learning curves - Recency bias weighting: recent 7 days weighted 3x higher (exponential decay) - Confidence scoring prevents overfitting: min(1.0, executions / 20) - Learning curves computed from daily execution windows ### Agent Scoring Service (✅ Complete) - Unified AgentScore combining SwarmCoordinator + learning profiles - Scoring formula: 0.3*base + 0.5*expertise + 0.2*confidence - Rank agents by combined score for intelligent assignment - Support for recency-biased scoring (recent_success_rate) - Methods: rank_agents, select_best, rank_agents_with_recency ### KG Integration (✅ Complete) - KGPersistence::get_executions_for_task_type() - query by agent + task type - KGPersistence::get_agent_executions() - all executions for agent - Coordinator::load_learning_profile_from_kg() - core KG→Learning integration - Coordinator::load_all_learning_profiles() - batch load for multiple agents - Convert PersistedExecution → ExecutionData for learning calculations ### Agent Assignment Integration (✅ Complete) - AgentCoordinator uses learning profiles for task assignment - extract_task_type() infers task type from title/description - assign_task() scores candidates using AgentScoringService - Fallback to load-based selection if no learning data available - Learning profiles stored in coordinator.learning_profiles RwLock ### Profile Adapter Enhancements (✅ Complete) - create_learning_profile() - initialize empty profiles - add_task_type_expertise() - set task-type expertise - update_profile_with_learning() - update swarm profiles from learning ## Files Modified ### vapora-knowledge-graph/src/persistence.rs (+30 lines) - get_executions_for_task_type(agent_id, task_type, limit) - get_agent_executions(agent_id, limit) ### vapora-agents/src/coordinator.rs (+100 lines) - load_learning_profile_from_kg() - core KG integration method - load_all_learning_profiles() - batch loading for agents - assign_task() already uses learning-based scoring via AgentScoringService ### Existing Complete Implementation - vapora-knowledge-graph/src/learning.rs - calculation functions - vapora-agents/src/learning_profile.rs - data structures and expertise - vapora-agents/src/scoring.rs - unified scoring service - vapora-agents/src/profile_adapter.rs - adapter methods ## Tests Passing - learning_profile: 7 tests ✅ - scoring: 5 tests ✅ - profile_adapter: 6 tests ✅ - coordinator: learning-specific tests ✅ ## Data Flow 1. Task arrives → AgentCoordinator::assign_task() 2. Extract task_type from description 3. Query KG for task-type executions (load_learning_profile_from_kg) 4. Calculate expertise with recency bias 5. Score candidates (SwarmCoordinator + learning) 6. Assign to top-scored agent 7. Execution result → KG → Update learning profiles ## Key Design Decisions ✅ Recency bias: 7-day half-life with 3x weight for recent performance ✅ Confidence scoring: min(1.0, total_executions / 20) prevents overfitting ✅ Hierarchical scoring: 30% base load, 50% expertise, 20% confidence ✅ KG query limit: 100 recent executions per task-type for performance ✅ Async loading: load_learning_profile_from_kg supports concurrent loads ## Next: Phase 5.4 - Cost Optimization Ready to implement budget enforcement and cost-aware provider selection.
2026-01-11 13:03:53 +00:00
#!/usr/bin/env nu
# VAPORA Kubernetes Deployment Script
# Deploys VAPORA v2.0 to Kubernetes cluster
def main [
--namespace: string = "vapora" # Kubernetes namespace
--registry: string = "docker.io" # Docker registry
--skip-secrets # Skip secrets creation (if already exists)
--dry-run # Perform dry run without actual deployment
] {
print $"(ansi green)🚀 VAPORA K8s Deployment Script(ansi reset)"
print $"(ansi blue)═══════════════════════════════════════════════(ansi reset)"
print $"Namespace: ($namespace)"
print $"Registry: ($registry)"
print ""
# Check prerequisites
print $"(ansi yellow)📋 Checking prerequisites...(ansi reset)"
check_prerequisites
# Create namespace
print ""
print $"(ansi yellow)📦 Creating namespace...(ansi reset)"
if $dry_run {
kubectl create namespace $namespace --dry-run=client -o yaml
} else {
kubectl create namespace $namespace --dry-run=client -o yaml | kubectl apply -f -
}
# Create secrets (if not skipped)
if not $skip_secrets {
print ""
print $"(ansi yellow)🔐 Creating secrets...(ansi reset)"
print $"(ansi red)⚠️ WARNING: Update secrets in kubernetes/03-secrets.yaml before production deployment!(ansi reset)"
if $dry_run {
kubectl apply -f kubernetes/03-secrets.yaml --dry-run=client
} else {
kubectl apply -f kubernetes/03-secrets.yaml
}
}
# Apply manifests in order
print ""
print $"(ansi yellow)📝 Applying Kubernetes manifests...(ansi reset)"
let manifests = [
"kubernetes/00-namespace.yaml"
"kubernetes/01-surrealdb.yaml"
"kubernetes/02-nats.yaml"
"kubernetes/04-backend.yaml"
"kubernetes/05-frontend.yaml"
"kubernetes/06-agents.yaml"
"kubernetes/07-mcp-server.yaml"
"kubernetes/08-ingress.yaml"
]
for manifest in $manifests {
print $" (ansi cyan)Applying ($manifest)...(ansi reset)"
if $dry_run {
kubectl apply -f $manifest --dry-run=client
} else {
kubectl apply -f $manifest
}
}
if not $dry_run {
# Wait for rollout
print ""
print $"(ansi yellow)⏳ Waiting for deployments to be ready...(ansi reset)"
try {
kubectl rollout status deployment/vapora-backend -n $namespace --timeout=5m
kubectl rollout status deployment/vapora-frontend -n $namespace --timeout=5m
kubectl rollout status deployment/vapora-agents -n $namespace --timeout=5m
kubectl rollout status deployment/vapora-mcp-server -n $namespace --timeout=5m
} catch {
print $"(ansi red)❌ Timeout waiting for deployments. Check status manually.(ansi reset)"
}
# Get status
print ""
print $"(ansi yellow)📊 Deployment status:(ansi reset)"
kubectl get all -n $namespace
print ""
print $"(ansi yellow)🌐 Ingress endpoints:(ansi reset)"
kubectl get ingress -n $namespace
print ""
print $"(ansi green)✅ Deployment complete!(ansi reset)"
print ""
print $"(ansi cyan)Next steps:(ansi reset)"
print " 1. Update ingress hostname in kubernetes/08-ingress.yaml"
print " 2. Configure DNS to point to ingress IP"
print " 3. Access UI at configured domain"
print " 4. Monitor logs: kubectl logs -n vapora -l app=vapora-backend"
} else {
print ""
print $"(ansi green)✅ Dry run complete! No changes were made.(ansi reset)"
}
}
# Check if required tools are installed
def check_prerequisites [] {
let required_tools = ["kubectl"]
for tool in $required_tools {
if (which $tool | is-empty) {
print $"(ansi red)❌ Error: ($tool) is not installed(ansi reset)"
exit 1
}
}
# Check kubectl cluster connection
try {
kubectl cluster-info | ignore
print $"(ansi green)✅ kubectl configured and connected(ansi reset)"
} catch {
print $"(ansi red)❌ Error: kubectl not configured or cluster not accessible(ansi reset)"
exit 1
}
# Check if kubernetes manifests exist
if not ("kubernetes" | path exists) {
print $"(ansi red)❌ Error: kubernetes/ directory not found(ansi reset)"
exit 1
}
}