performance-analysis

Analyze system performance metrics and profiling data to identify bottlenecks and prioritize optimizations.

1|1|Updated May 16, 2026

Search Tags:#monitoring #optimization #migration #profiling #latency #bottleneck detection #performance analysis

Ready to run

npx skills add https://github.com/vanduc2514/hackathon-lablab-ibm-bob --skill performance-analysis-vanduc2514

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: performance-analysis
Source: https://github.com/vanduc2514/hackathon-lablab-ibm-bob/tree/main/.bob/skills/performance-analysis
Command: npx skills add https://github.com/vanduc2514/hackathon-lablab-ibm-bob --skill performance-analysis-vanduc2514

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? After a migration, teams often lack a clear picture of how the new system performs under real load, making it hard to find bottlenecks and decide which optimizations to tackle first. ## Core Features & Use Cases - Baseline Measurement: Collects throughput, latency percentiles, resource utilization, and error rates from production or staging environments. - Bottleneck Identification: Profiles CPU hotspots, memory allocation, I/O wait times, and database query performance, ranking issues by impact. - Prioritized Optimization Roadmap: Produces an `optimization_performance_analysis.md` report comparing current metrics against SLA targets with ranked recommendations. - Use Case: After migrating a Spring Boot application, run this analysis to discover an N+1 query pattern adding 400ms to P95 latency, then receive a prioritized fix list ordered by expected impact versus effort. ## Quick Start Analyze the performance of my migrated application using the validation report and production metrics, and create a prioritized optimization analysis document.

Frequently Asked Questions about performance-analysis

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I identify performance bottlenecks after a migration?▼

Collect baseline metrics from production or staging, then profile the application under realistic load to find CPU, memory, I/O, and database hotspots. Rank each bottleneck by its quantified impact on latency and throughput to build a prioritized fix list.

What profiling tools should I use for Java or Node.js applications?▼

For Java, use JProfiler, YourKit, VisualVM, async-profiler, or Java Flight Recorder. For Node.js, use Chrome DevTools, clinic.js, 0x, or node --prof. Python supports cProfile and py-spy, while .NET works with dotTrace and PerfView.

Which performance metrics matter most for SLA comparison?▼

Focus on throughput (requests per second), latency percentiles (P50, P95, P99), resource utilization (CPU, memory, disk, network), and error rates. Database query times and external dependency latency also directly affect SLA compliance.

Why should performance testing use production-like load instead of benchmarks?▼

Synthetic benchmarks rarely reproduce real traffic patterns, so they miss bottlenecks like N+1 queries, connection pool exhaustion, or slow external API calls. Profiling under realistic load yields data that reflects actual user experience.

What inputs are required before running a performance analysis?▼

You need a validation report confirming functional correctness, design documents describing the target architecture, access to production or staging metrics and profiling tools, and defined performance targets such as SLAs and latency goals.