commands/logging/monitor-system-resources.md

Monitor System Resources

You are helping the user monitor live system resource usage (CPU, RAM, disk I/O, network).

Task

1. Quick resource overview:

   # System overview
   top -b -n 1 | head -20
   
   # Better overview with htop (if installed)
   htop
   
   # Modern alternative: btop
   btop
   

2. CPU monitoring:

   # CPU usage summary
   mpstat 1 5  # 5 samples, 1 second interval
   
   # Per-core usage
   mpstat -P ALL 1 5
   
   # Top CPU consumers
   ps aux --sort=-%cpu | head -10
   
   # CPU frequency and temperature
   watch -n 1 "grep MHz /proc/cpuinfo | head -20 && sensors"
   

3. Memory monitoring:

   # Memory usage
   free -h
   
   # Detailed memory info
   cat /proc/meminfo
   
   # Top memory consumers
   ps aux --sort=-%mem | head -10
   
   # Memory usage by process
   smem -tk  # If smem is installed
   
   # Watch memory usage
   watch -n 1 free -h
   

4. Disk I/O monitoring:

   # I/O statistics
   iostat -x 1 5
   
   # Disk usage by device
   iotop -o  # Only show active I/O
   
   # Per-process I/O
   iotop -P
   
   # Watch disk I/O
   watch -n 1 "iostat -x | grep -E 'Device|sd|nvme'"
   

5. Network monitoring:

   # Network interface statistics
   ifstat 1 5
   
   # Bandwidth per interface
   bmon
   
   # Network speed
   iftop
   
   # Connection summary
   ss -s
   
   # Active connections
   nethogs
   

6. Combined system monitoring:

   # All-in-one monitoring
   glances
   
   # Custom dashboard
   watch -n 1 '
   echo "=== CPU ==="
   top -b -n 1 | head -5 | tail -2
   echo ""
   echo "=== Memory ==="
   free -h | grep -E "Mem|Swap"
   echo ""
   echo "=== Disk I/O ==="
   iostat -x 1 1 | grep -E "Device|sd|nvme" | head -5
   echo ""
   echo "=== Network ==="
   ifstat 1 1 | tail -1
   '
   

7. GPU monitoring (AMD):

   # AMD GPU usage
   radeontop
   
   # GPU sensors
   watch -n 1 "rocm-smi"
   
   # Detailed GPU info
   watch -n 1 "rocm-smi --showuse --showmemuse --showtemp"
   

8. Process tree with resource usage:

   # Process tree
   pstree -p
   
   # Resource-aware process tree
   ps auxf
   
   # Find resource hogs
   ps -eo pid,ppid,cmd,%mem,%cpu --sort=-%cpu | head -20
   

9. System load monitoring:

   # Load average
   uptime
   
   # Load over time
   watch -n 1 "uptime && cat /proc/loadavg"
   
   # Who's causing load
   tload -d 1
   

10. Create monitoring script:

    cat > /tmp/system-monitor.sh << 'EOF'
    #!/bin/bash
    echo "System Resource Monitor - $(date)"
    echo "========================================"
    echo ""
    echo "CPU Usage:"
    mpstat 1 1 | tail -1
    echo ""
    echo "Memory:"
    free -h | grep -E "Mem|Swap"
    echo ""
    echo "Load Average:"
    uptime
    echo ""
    echo "Top 5 CPU Processes:"
    ps aux --sort=-%cpu | head -6 | tail -5
    echo ""
    echo "Top 5 Memory Processes:"
    ps aux --sort=-%mem | head -6 | tail -5
    echo ""
    echo "Disk Usage:"
    df -h / /home
    echo ""
    echo "Disk I/O:"
    iostat -x 1 1 | grep -E "sd|nvme"
    EOF
    
    chmod +x /tmp/system-monitor.sh
    /tmp/system-monitor.sh
    

Present Summary to User

Provide snapshot of:

• CPU: Usage %, load average, top processes
• Memory: Used/Free, swap usage, top consumers
• Disk: Usage %, I/O wait, active reads/writes
• Network: Bandwidth usage, active connections
• GPU: Usage % (if applicable)

Flag any concerning patterns:

• High CPU usage (>80% sustained)
• Low memory (<10% free)
• High swap usage
• Disk I/O bottlenecks
• Network saturation

Install Monitoring Tools if Needed

# Install commonly needed tools
sudo apt install -y \
    htop \
    btop \
    iotop \
    iftop \
    nethogs \
    glances \
    sysstat \
    bmon \
    radeontop

Troubleshooting High Usage

High CPU:

• Identify process: top or htop
• Check if legitimate (updates, backups, encoding)
• Kill if necessary: kill -15 PID

High Memory:

• Check for memory leaks: ps aux --sort=-%mem
• Clear caches: sudo sync && sudo sysctl -w vm.drop_caches=3
• Check for swap thrashing

High Disk I/O:

• Identify process: iotop -o
• Check if expected (backups, indexing, updates)
• Monitor disk health: /check-disk-errors

High Network:

• Identify connections: nethogs or iftop
• Check for unexpected traffic
• Use ss -tunap to see connections

Notes

• Most monitoring tools require sudo for full functionality
• glances provides best all-in-one view
• btop is modern, colorful alternative to htop
• Some tools need to be installed separately
• For persistent monitoring, consider setting up Prometheus/Grafana
• Check systemd-cgtop for cgroup resource usage