performance

Performance

A structured guide to profiling, caching, database optimization, async patterns, and performance budgets for production services.

When to Activate

• Profiling a slow endpoint or service
• Implementing a caching layer (in-process, Redis, or HTTP)
• Optimizing a database query or fixing N+1 problems
• Setting a performance budget for an API endpoint
• Reducing memory usage or GC pressure
• Choosing between sync and async patterns for a workload

---

Profiling

When to Profile

• Profile before optimizing — never guess where the bottleneck is
• CPU profiling — where is time spent (function call time)?
• Memory profiling — what objects are consuming heap space?
• I/O profiling — what is blocking on disk or network?

Python — cProfile + snakeviz

import cProfile
import pstats
import io

pr = cProfile.Profile()
pr.enable()
result = my_slow_function()
pr.disable()

s = io.StringIO()
ps = pstats.Stats(pr, stream=s).sort_stats('cumulative')
ps.print_stats(20)  # top 20 slowest functions
print(s.getvalue())

# Profile a whole script from the command line:
# python -m cProfile -o output.prof script.py
# snakeviz output.prof  # opens interactive flame graph in browser

Memory profiling with memory_profiler:

# pip install memory-profiler
from memory_profiler import profile

@profile
def my_function():
    # annotated line-by-line memory usage
    data = [x for x in range(10_000_000)]
    return data

TypeScript/Node.js — clinic.js + 0x

# CPU flame graph
npx 0x -- node dist/server.js
# Opens a generated .html flame graph in the browser

# Heap snapshot + event loop lag
npx clinic doctor -- node dist/server.js

# CPU flame graph via clinic
npx clinic flame -- node dist/server.js

# Async waterfall / I/O bottlenecks
npx clinic bubbleprof -- node dist/server.js

Go — pprof

import (
    "net/http"
    _ "net/http/pprof" // side-effect import registers /debug/pprof handlers
)

// In main(), run alongside your app server:
go func() {
    http.ListenAndServe("localhost:6060", nil)
}()