How to Linux performance monitoring

Quick Health Snapshot

# uptime, load averages, logged-in users
uptime

# memory usage (human readable)
free -h

# disk usage by filesystem
df -h

CPU Usage

# interactive overview
top
# improved interface (install first)
sudo apt install -y htop || sudo dnf install -y htop
htop

# per-core stats every 2s (sysstat package)
# Ubuntu/Debian:
sudo apt install -y sysstat
# CentOS/AlmaLinux:
sudo dnf install -y sysstat
mpstat -P ALL 2

Tip: In top/htop, press P to sort by CPU.

Memory & Swap

# processes + memory over time
vmstat 2 10

# who is using swap?
grep -H 'VmSwap' /proc/*/status 2>/dev/null | sed 's/\/proc\///;s/\/status:/ /' | head

Tip: High swap use with low RAM free indicates memory pressure.

Disk I/O (Reads/Writes)

# top-like view of I/O per process
sudo apt install -y iotop || sudo dnf install -y iotop
sudo iotop

# device throughput and latency
iostat -xz 5 3   # from sysstat

Watch for high await/svctm and busy (%util) disks.

Filesystem Usage & Heavy Folders

# usage by mount
df -h

# biggest directories (top 20) under /
sudo du -xh / | sort -h | tail -n 20

# focused scan (fast)
sudo du -xh /var | sort -h | tail -n 20

Network Throughput

# quick interface stats
ip -s link

# per-connection overview (modern replacement for netstat)
ss -tulpn

System Logs & Alerts

# live systemd journal
sudo journalctl -f

# kernel only
sudo journalctl -k -f

Investigate services that spike CPU/I/O using journalctl -u <service>.

Create a Baseline & Automate Checks

1. Record normal ranges for load, memory, and disk I/O.
2. Schedule a daily report:

cat <<'SH' | sudo tee /usr/local/bin/sys-health.sh
#!/usr/bin/env bash
echo "=== $(date) ==="
uptime
free -h
df -h
iostat -xz 1 2 | tail -n +4
SH
sudo chmod +x /usr/local/bin/sys-health.sh
(crontab -l; echo "0 8 * * * /usr/local/bin/sys-health.sh >> ~/sys-health.log") | crontab -