Announcing MCP Toolbox Java SDK v1.0: Agentic data access for the enterprise
cloud.google.com·6d ago
Cloud & Infrastructure
The New Stack·15d ago·5 min read
Performance engineering from kernel analysis to AI: Adrian Cockcroft’s take
TL;DR
Adrian Cockcroft emphasizes the importance of analyzing response time distributions over traditional percentiles like P99, advocating for custom tools to better understand performance anomalies in modern web services.
✦ Why It Matters
Start analyzing response time distributions instead of relying solely on P99 metrics to uncover hidden performance issues.
Key Takeaways
- Explore response time distributions for deeper insights into performance.
- Develop custom tools to analyze performance anomalies effectively.
- Use statistical methods to identify multiple peaks in response times.
- Start performance analysis with a macro view before zooming in.
How It Works
Cockcroft's tool automates the identification of multiple peaks in response time distributions, allowing engineers to track fluctuations over time. By analyzing histograms rather than relying on averages, it provides a clearer picture of system performance.