performance-debug

mhalder 的头像作者 mhalder

诊断系统性能问题,包括 CPU、内存、磁盘和网络。当用户说“服务器速度慢”、“CPU 高”、“内存不足”、“磁盘已满”、“性能问题”或要求调试系统性能时使用。

📖 中文导读:诊断系统性能问题,包括 CPU、内存、磁盘和网络。当用户说“服务器速度慢”、“CPU 高”、“内存不足”、“磁盘已满”、“性能问题”或要求调试系统性能时使用。 以下正文为该技能的原始 SKILL.md 内容(英文),上方安装命令可直接使用。

name: performance-debug description: Diagnose system performance issues including CPU, memory, disk, and network. Use when the user says "server is slow", "high CPU", "out of memory", "disk full", "performance issues", or asks to debug system performance. allowed-tools: Bash, Read, Grep

Performance Debug

Diagnose CPU, memory, disk, and network performance bottlenecks.

Instructions

  1. Get system overview: top, htop, or vmstat
  2. Identify the bottleneck type (CPU, memory, disk, network)
  3. Drill down with specific tools
  4. Identify the root cause process/resource
  5. Recommend solutions

Quick overview

# System summary
uptime
free -h
df -h
top -bn1 | head -20

# All-in-one view
htop  # or top

CPU analysis

# High-level CPU usage
mpstat 1 5
top -bn1 -o %CPU | head -15

# Per-process CPU
pidstat 1 5
ps aux --sort=-%cpu | head -10

# Find CPU-intensive process
top -bn1 | grep -A10 "PID USER"

Memory analysis

# Memory overview
free -h
vmstat 1 5

# Per-process memory
ps aux --sort=-%mem | head -10
pidstat -r 1 5

# Memory details
cat /proc/meminfo
smem -tk  # if available

# Find memory leaks (growth over time)
watch -n 5 'ps aux --sort=-%mem | head -10'

Disk analysis

# Disk space
df -h
du -sh /* 2>/dev/null | sort -hr | head -10

# Disk I/O
iostat -x 1 5
iotop -b -n 5  # requires root

# Find large files
find / -type f -size +100M 2>/dev/null

# Find disk-heavy processes
pidstat -d 1 5

Network analysis

# Connections and bandwidth
ss -tuln
ss -tunap | grep ESTAB
nethogs  # per-process bandwidth

# Network statistics
netstat -i
ip -s link

# Check for connection issues
ping -c 4 8.8.8.8
mtr --report google.com

Common bottlenecks

Symptom Indicator Solution
Load > cores CPU-bound Identify hot process, scale horizontally
High %wa in top I/O wait Check disk, move to SSD, optimize queries
Low free + high swap Memory Find leak, increase RAM, tune OOM
High %si/%hi Interrupts NIC issue, driver problem

Rules

  • MUST check all four resources (CPU, memory, disk, network)
  • MUST identify specific processes causing issues
  • MUST provide actionable recommendations
  • Never kill processes without user approval
  • Always check if symptoms correlate with specific times (cron, traffic)