How AngusTester Defines Unified Test Metrics for Platform Businesses
Why Unified Test Metrics Are Necessary
Platform business and technical teams inherently have different perspectives on performance evaluation:
| Dimension | Business Team Focus | Technical Team Focus | Unified Metric Value |
|---|---|---|---|
| Performance | User Experience/SLA Compliance | TPS/Latency/Error Rate | Align business goals with technical implementation |
| Functionality | Core Process Coverage | Branch Coverage/Code Quality | Ensure functional completeness and technical rigor |
| Stability | System Availability | Error Recovery Time/Resource Redundancy | Balance availability requirements with system robustness |
Core Value of Unified Metrics
Value Realization Matrix
🚀 Consistency Guarantee
- Cross-team comparable test results
- Visualized performance trends across versions
- Eliminate metric interpretation ambiguities
📊 Standardized Quality Assessment
- Establish SLA compliance evaluation system
- Quantitative analysis of functional coverage
- Unified defect density measurement
🔍 Problem Tracking Efficiency
- Automatic alerts for metric anomalies
- 50%+ reduction in root cause identification time
- Optimized MTTR (Mean Time To Repair)
🔄 Continuous Improvement Driver
- Establish performance baseline monitoring
- Identify key optimization points
- Quantitatively verify improvement effectiveness
Platform-Level Standardized Configuration
Global Metric Setup Process
Configuration Steps
- Navigate to
Configuration → Metrics - Modify default metric thresholds:
- Transactions per second (TPS)
- Concurrency (Users)
- Response Time (P90/P95/P99)
- Error Rate Upper Limit
- Resource Utilization Threshold
- Save settings to take effect platform-wide
Recommended Platform Baseline Metrics
| Metric Type | Recommended Value | Applicable Scenario |
|---|---|---|
| Response Time (P90) | ≤500ms | User-facing interactive systems |
| Error Rate | ≤0.1% | Core business-critical systems |
| CPU Peak | ≤75% | Compute-intensive services |
| Memory Usage | ≤70% | Memory-sensitive systems |
API-Level Customization Solution
Dedicated Metric Configuration Process
Configuration Steps
- Access API management:
APIs → Target Service → Target API - Navigate to
Test Metricstab - Configure API-specific metrics:
- Required transactions per second (TPS)
- Required concurrency (users)
- Special response time requirements
- Error tolerance for specific scenarios
- Custom resource usage rules
- Save to take effect only for this API
Custom Metric Application Scenarios
| Scenario Type | Custom Metric Recommendations | Business Value |
|---|---|---|
| Core Payment APIs | Response time≤200ms Error rate=0% | Guarantee transaction success rate |
| Big Data Analysis APIs | Response time≤5s Allow partial timeouts | Balance performance and data integrity |
| Scheduled Batch Processing | CPU peak≤85% Allow instant high load | Maximize resource utilization |
| External Dependency APIs | Error rate≤5% Timeout rate≤1% | Compatibility with unstable third-party services |
Unify Your Test Metrics System Now: 🔗 Access AngusTester Console 🔗