Deploy Checklist

Pre-Deploy Validation

Environment Variables

# Verify all required env vars are set
REQUIRED_VARS=(
    "DATABASE_URL"
    "SECRET_KEY"
    "REDIS_URL"
    "API_KEY"
    "LOG_LEVEL"
)

for var in "${REQUIRED_VARS[@]}"; do
    if [ -z "${!var}" ]; then
        echo "ERROR: Missing required env var: $var"
        exit 1
    fi
done

# Check for hardcoded secrets
if grep -r "password\s*=\s*['\"][^'\"]\{8,\}" src/ --include="*.py" --include="*.js"; then
    echo "ERROR: Potential hardcoded password found"
    exit 1
fi

Dependency Check

# Python
pip install -r requirements.txt --dry-run
# or
pip check

# Node
npm ci --prefer-offline
npm audit --audit-level=high

# Go
go mod download
go mod verify

# Docker image layers
docker build --progress=plain -t app:deploy .

Migration Safety

# Django
python manage.py migrate --check
python manage.py sqlmigrate app 0002 --backwards

# Rails
rake db:migrate:status
rake db:rollback STEP=1  # test rollback path

# PostgreSQL (general)
psql -c "SELECT * FROM schema_migrations ORDER BY version DESC LIMIT 5;"

Deployment Command

# Blue-green deployment
./deploy.sh --strategy blue-green --slot green

# Canary deployment
./deploy.sh --strategy canary --weight 10%

# Rolling deployment
./deploy.sh --strategy rolling --max-surge=1 --max-unavailable=0

Rollback Plan

#!/bin/bash
# rollback.sh

VERSION=${1:-$(cat VERSION.prev)}
echo "Rolling back to version: $VERSION"

# 1. Stop new deployment
kubectl rollout undo deployment/app

# 2. Verify rollback
kubectl rollout status deployment/app --timeout=120s

# 3. Run health check
curl -f http://localhost:8080/health || exit 1

# 4. Verify database compatibility
python manage.py migrate --check

# 5. Notify team
echo "Rollback to $VERSION complete"

Post-Deploy Verification

# Health check
curl -f http://localhost:8080/health
curl -f http://localhost:8080/ready

# Smoke tests
pytest tests/smoke/ -v

# Verify key endpoints
for path in /api/users /api/orders /api/health; do
    response=$(curl -s -o /dev/null -w "%{http_code}" http://localhost:8080$path)
    if [ "$response" != "200" ]; then
        echo "FAIL: $path returned $response"
        exit 1
    fi
done

# Check error rate
errors=$(curl -s http://localhost:9090/api/errors | jq '.count')
if [ "$errors" -gt 10 ]; then
    echo "WARN: Elevated error count: $errors"
fi

# Verify database connections
psql -c "SELECT count(*) FROM pg_stat_activity WHERE state = 'active';"

Deployment Checklist

Before Deploy

• All tests passing (unit, integration, smoke)
• No pending database migrations that are backwards-incompatible
• Environment variables verified
• Secrets rotated if deployed to new environment
• Rollback plan documented and tested
• Team notified (Slack/DM)
• Monitoring dashboards ready

During Deploy

• Deployment strategy selected (rolling/blue-green/canary)
• Progress monitored in real-time
• Error rate tracked
• Latency monitored
• Rollback trigger defined (error rate > 1% for 5min)

After Deploy

• Health checks pass
• Key user flows verified (smoke tests)
• Error rate within baseline
• Database migrations applied successfully
• Performance benchmarks within threshold
• Rollback version recorded for future reference
• Deployment log archived

Rollback Triggers

Metric	Threshold	Action
Error rate	> 1% for 5 min	Auto-rollback
p99 latency	> 2x baseline	Alert, prepare rollback
Database connection pool	90% utilized	Slow down deployment
Disk usage	> 85%	Halt deployment
Out-of-memory errors	Any	Immediate rollback

Configuration Template

# deploy-config.yaml
version: "2.0"
strategy: blue-green
health_check:
  path: /health
  interval: 10s
  timeout: 5s
  healthy_threshold: 3
  unhealthy_threshold: 2
rollback:
  auto: true
  triggers:
    - metric: error_rate
      threshold: 0.01
      duration: 5m
    - metric: p99_latency_ms
      threshold: 1000
      duration: 3m
notifications:
  on_deploy: "#deployments"
  on_rollback: "#deployments"
  on_success: "#deployments"