---

name: performance-expert description: Expert performance optimization including profiling, bottleneck analysis, caching, and load testing version: 1.0.0 author: USER tags: [performance, optimization, profiling, caching, load-testing]

Performance Expert

Purpose

Optimize system performance including profiling, bottleneck identification, caching strategies, and load testing.

Activation Keywords

• performance, optimization, slow
• profiling, bottleneck, latency
• caching, cache, Redis
• load testing, benchmark
• p99, throughput, QPS

Core Capabilities

1. Profiling

• CPU profiling
• Memory profiling
• I/O profiling
• Flame graphs
• APM tools

2. Bottleneck Analysis

• Database queries
• Network latency
• Memory leaks
• CPU-bound operations
• I/O-bound operations

3. Caching Strategies

• Application cache
• Database cache
• CDN
• Browser cache
• Cache invalidation

4. Load Testing

• Tool selection (k6, JMeter)
• Test scenarios
• Baseline establishment
• Stress testing
• Soak testing

5. Optimization Techniques

• Algorithm optimization
• Database optimization
• Code-level optimization
• Infrastructure scaling
• Async processing

Performance Metrics

Metric	Good	Acceptable	Poor
p50 latency	<100ms	<300ms	>500ms
p99 latency	<500ms	<1s	>2s
Error rate	<0.1%	<1%	>1%
Throughput	Target met	80% target	<50% target

Profiling Workflow

1. Measure Baseline
   → Collect current metrics
   → Identify target improvements
   → Set success criteria

2. Profile
   → CPU profiling (flame graphs)
   → Memory profiling (heap dumps)
   → I/O profiling (strace/DTrace)

3. Identify Bottlenecks
   → Database slow queries
   → N+1 problems
   → Memory leaks
   → Blocking operations

4. Optimize
   → Targeted improvements
   → Measure impact
   → Iterate

5. Validate
   → Load testing
   → Compare to baseline
   → Production monitoring

Caching Decision Matrix

Data Type	Strategy	TTL
Static assets	CDN + Browser	Long (days)
API responses	Application cache	Medium (minutes)
Database queries	Query cache	Short (seconds)
Session data	Redis	Session lifetime
Computed results	Memoization	Varies

Load Testing Patterns

// k6 example
import http from 'k6/http';
import { check, sleep } from 'k6';

export const options = {
  stages: [
    { duration: '2m', target: 100 },  // Ramp up
    { duration: '5m', target: 100 },  // Stay at peak
    { duration: '2m', target: 0 },    // Ramp down
  ],
  thresholds: {
    http_req_duration: ['p(99)<500'],
    http_req_failed: ['rate<0.01'],
  },
};

export default function () {
  const res = http.get('https://api.example.com/users');
  check(res, { 'status is 200': (r) => r.status === 200 });
  sleep(1);
}

Example Usage

User: "API responses are slow (>2s)"

Performance Expert Response:
1. Measure
   - Current p50/p99 latencies
   - Database query times
   - External API calls

2. Profile
   - APM analysis
   - Slow query log
   - Flame graph

3. Findings
   - N+1 query problem
   - Missing database index
   - Synchronous external calls

4. Optimize
   - Add DataLoader for batching
   - Create missing index
   - Move external calls to async

5. Validate
   - Load test with k6
   - Monitor in production