HeliosDB Production Deployment Guide¶

Overview¶

Comprehensive guide for deploying HeliosDB to production using automated infrastructure.

Prerequisites¶

Terraform >= 1.5.0
kubectl >= 1.28.0
Helm >= 3.13.0
AWS CLI >= 2.13.0
Docker >= 24.0.0

Quick Start¶

# 1. Deploy infrastructure
cd infrastructure/terraform
terraform apply -var-file="environments/production.tfvars"

# 2. Configure kubectl
aws eks update-kubeconfig --name heliosdb-production-cluster --region us-west-2

# 3. Deploy HeliosDB
./scripts/deploy/deploy.sh --environment production --version 7.0.1 --strategy rolling

Deployment Strategies¶

Rolling (Default) - For minor updates¶

Zero-downtime gradual replacement
Automatic rollback on failure

Blue-Green - For major updates¶

Deploy to inactive, switch traffic
Requires 2x resources

Canary - For risky changes¶

Progressive: 10% → 50% → 100%
Early issue detection

Operations¶

Deploy¶

./scripts/deploy/deploy.sh --environment production --version 7.0.1 --strategy canary

Rollback¶

./scripts/deploy/rollback.sh --environment production

Health Check¶

./scripts/deploy/health-check.sh --namespace heliosdb

Scale¶

./scripts/deploy/scale.sh --replicas 7
./scripts/deploy/scale.sh --enable-hpa --min 5 --max 15 --cpu-target 70

Monitoring¶

Prometheus: Port 9090
Grafana: kubectl port-forward -n monitoring svc/prometheus-grafana 3000:80
Health: https://<endpoint>:8443/health

Troubleshooting¶

# Check pods
kubectl get pods -n heliosdb
kubectl logs -n heliosdb <pod-name>

# Check resources
kubectl top pods -n heliosdb
kubectl describe pod -n heliosdb <pod-name>

# Check services
kubectl get svc,endpoints,networkpolicy -n heliosdb

Support¶

Ops: ops-oncall@example.com
Slack: #heliosdb-ops