Deployment Health Check
Validates deployment success through comprehensive health monitoring and regression detection.
When invoked directly with /datadops:deployment-health-check, use $ARGUMENTS as the deployment context.
If needed, ask for the service name, version, environment, and deployment window before continuing.
Capabilities
📊 Before/After Comparison
- Baseline metrics capture pre-deployment
- Post-deployment performance analysis
- Statistical significance testing
- Regression severity assessment
🔍 Multi-Dimensional Health Check
- Golden Signals: Error rate, latency, throughput, saturation
- Custom Metrics: Business KPIs and service-specific indicators
- Log Patterns: New error types, warning frequencies
- APM Performance: Trace latency distribution, dependency health
⚡ Early Warning System
- Real-time monitoring during deployment window
- Automated anomaly detection
- Progressive rollout validation
- Canary deployment analysis
Health Check Dimensions
1. Service Performance
- Request success rate (target: > 99.5%)
- Response time percentiles (P50, P95, P99)
- Throughput consistency
- Memory/CPU utilization patterns
2. Error Patterns
- New error types or frequencies
- HTTP status code distributions
- Exception rate changes
- Dependency failure impacts
3. Infrastructure Health
- Host resource utilization
- Container restart rates
- Network connectivity
- Database connection pools
4. Business Metrics
- Conversion rates
- User engagement metrics
- Revenue-impacting indicators
- Customer experience scores
Workflow Process
Phase 1: Baseline Capture (Pre-deployment)
-
Metrics Snapshot
- Capture 1-hour baseline before deployment
- Record key performance indicators
- Document expected traffic patterns
-
Health Thresholds
- Calculate statistical baselines
- Set regression detection thresholds
- Define success/failure criteria
Phase 2: Deployment Monitoring (During)
- Real-time Tracking
- Monitor deployment events and progress
- Track immediate health indicators
- Alert on deployment-blocking issues
Phase 3: Post-deployment Validation (After)
-
Comparison Analysis
- Compare 30-60 minutes post vs baseline
- Statistical significance testing
- Trend analysis and projections
-
Health Assessment
- Overall deployment health score
- Specific regression identification
- Risk assessment and recommendations
Usage Examples
Standard Deployment Check
"I just deployed version 2.1.4 of the payment service. Validate the deployment health."
Expected Response:
- Deployment event correlation
- Before/after metric comparisons
- Error rate and latency analysis
- Health score and recommendations
Canary Deployment Validation
"Check if the canary deployment of checkout service is performing well compared to production."
Expected Response:
- Canary vs production performance comparison
- Statistical significance of differences
- Recommendations for traffic ramp-up
- Rollback triggers if needed
Rollback Decision Support
"Should we rollback the API service deployment? Performance seems degraded."
Expected Response:
- Quantified performance regression analysis
- Impact assessment (user experience, business metrics)
- Rollback recommendation with confidence level
- Monitoring plan if continuing with deployment
Health Check Templates
Microservice Health Check
- API response times and success rates
- Database query performance
- Cache hit rates and performance
- Downstream service dependencies
Frontend Application Check
- Page load times and user interactions
- JavaScript error rates
- CDN performance metrics
- User engagement indicators
Infrastructure Service Check
- Resource utilization patterns
- Service discovery health
- Load balancer performance
- Auto-scaling behavior
Integration Patterns
CI/CD Pipeline Integration
# Claude Code interactive invocation
/datadops:deployment-health-check payment service version 2.1.4 in production
# Prompt-based automation example
claude -p "Validate deployment health for $SERVICE_NAME version $GITHUB_SHA in $ENVIRONMENT. Compare pre-deployment and post-deployment metrics, logs, spans, and events, then give a go/no-go recommendation."
Automated Rollback Triggers
- Health score below threshold (< 80%)
- Error rate increase > 2x baseline
- Latency regression > 50% increase
- Critical business metric degradation
Risk Assessment Framework
Low Risk (Score: 90-100%)
- All metrics within expected ranges
- No significant error pattern changes
- Performance improvements or neutral
- Action: Continue monitoring
Medium Risk (Score: 70-89%)
- Minor performance regressions
- Slight error rate increases
- Some concerning log patterns
- Action: Enhanced monitoring, prepare rollback
High Risk (Score: < 70%)
- Significant performance degradation
- Major error rate increases
- Customer-impacting issues
- Action: Consider immediate rollback
Success Criteria
- Detection Speed: Identify regressions within 15 minutes
- Accuracy: < 5% false positive rate on health assessments
- Coverage: Monitor 100% of critical service indicators
- Actionability: Clear go/no-go recommendations with confidence scores
Customization Options
Service-Specific Thresholds
- Custom success rate targets per service
- Business-critical metric weightings
- Environment-specific baselines
Deployment Window Awareness
- Time-based baseline adjustments
- Traffic pattern expectations
- Seasonal/cyclical considerations