Docker Performance Optimization 🚀

╔════════════════════════════════════════════════════╗
║  DOCKER PERFORMANCE OPTIMIZATION                   ║
║                                                   ║
║  CPU | Memory | I/O | Network | Build | Runtime   ║
╚════════════════════════════════════════════════════╝

Resource Management 📊

CPU Optimization

# Limitar CPU
docker run --cpus=".5" nginx

# Definir CPU shares
docker run --cpu-shares=512 nginx

# Pinning em CPUs específicas
docker run --cpuset-cpus="0,1" nginx

Memory Management

# Limitar memória
docker run --memory="512m" nginx

# Configurar swap
docker run --memory-swap="1g" nginx

# Reservar memória
docker run --memory-reservation="256m" nginx

Storage Performance 💾

Storage Driver Selection

{
  "storage-driver": "overlay2",
  "storage-opts": [
    "overlay2.override_kernel_check=true"
  ]
}

Volume Performance

# Volume local otimizado
docker volume create --driver local \
  --opt type=tmpfs \
  --opt device=tmpfs \
  --opt o=size=1g \
  fast-volume

# Monitorar I/O
docker stats --format "table {{.Name}}\t{{.BlockIO}}"

Network Optimization 🌐

Network Mode Selection

# Host networking
docker run --network host nginx

# Performance tuning
docker run \
  --network-alias=fastnet \
  --dns=8.8.8.8 \
  nginx

Network Metrics

# Monitorar tráfego
docker stats --format "table {{.Name}}\t{{.NetIO}}"

# Análise detalhada
docker run --rm \
  --net container:target-container \
  nicolaka/netshoot iftop

Build Performance ⚡

Multi-stage Builds

# Build stage
FROM golang:1.19 AS builder
WORKDIR /app
COPY . .
RUN CGO_ENABLED=0 go build -o app

# Production stage
FROM alpine:3.14
COPY --from=builder /app/app /
CMD ["/app"]

Cache Optimization

# Usar buildkit
export DOCKER_BUILDKIT=1

# Cache em registry
docker buildx build \
  --cache-from type=registry,ref=user/app:cache \
  --cache-to type=registry,ref=user/app:cache \
  .

Runtime Performance 🔄

Container Stats

# Métricas em tempo real
docker stats --no-stream

# Formato personalizado
docker stats --format \
  "table {{.Container}}\t{{.CPUPerc}}\t{{.MemUsage}}\t{{.NetIO}}"

Resource Quotas

version: '3.8'
services:
  app:
    deploy:
      resources:
        limits:
          cpus: '0.50'
          memory: 512M
        reservations:
          cpus: '0.25'
          memory: 256M

Monitoring Tools 📈

Prometheus Integration

metrics:
  prometheus:
    enabled: true
    port: 9323

Grafana Dashboard

# Container metrics
rate(container_cpu_usage_seconds_total{container_name!=""}[5m])
sum(container_memory_usage_bytes) by (container_name)

Performance Testing 🔍

Load Testing

# Apache Benchmark
ab -n 1000 -c 10 http://localhost:8080/

# Hey HTTP load generator
hey -n 2000 -c 50 http://localhost:8080/

Profiling

# CPU profile
docker run --cap-add=SYS_PTRACE \
  --security-opt seccomp=unconfined \
  app perf record -F 99 -p 1

# Memory profile
docker run -p 6060:6060 app
go tool pprof http://localhost:6060/debug/pprof/heap

Best Practices Checklist ✅

Container Level

[x] Resource limits configurados
[x] Logging driver otimizado
[x] Network mode apropriado
[x] Volume mounts eficientes
[x] Image size otimizado

Host Level

[x] Storage driver adequado
[x] Sistema de arquivos otimizado
[x] Kernel parameters tuned
[x] Network stack otimizado
[x] Monitoramento ativo

Performance Troubleshooting Matrix 🎯

Sintoma	Métrica	Comando	Solução
CPU Alto	container_cpu_usage	`docker stats`	Ajustar limites
Memória Leak	container_memory_usage	`docker stats`	Heap analysis
I/O Lento	container_fs_io_current	`iostat`	Storage driver
Network Lag	container_network_transmit	`iftop`	Network mode

Advanced Tuning 🔧

Kernel Parameters

# Ajustar sysctls
docker run --sysctl net.core.somaxconn=1024 nginx

# Configurar ulimits
docker run --ulimit nofile=65535:65535 nginx

Runtime Flags

# Otimizar garbage collector
docker run -e "GOGC=100" app

# Tuning JVM
docker run -e "JAVA_OPTS=-XX:+UseG1GC" java-app

Waifu Performance Tips 💫

Quick Reference 📚

Common Commands

# CPU profile
docker stats --format "table {{.Container}}\t{{.CPUPerc}}"

# Memory analysis
docker stats --format "table {{.Container}}\t{{.MemUsage}}"

# I/O metrics
docker stats --format "table {{.Container}}\t{{.BlockIO}}"

# Network stats
docker stats --format "table {{.Container}}\t{{.NetIO}}"

Performance Tools

cAdvisor
Prometheus
Grafana
netdata
ctop

16 abril 2025