Essential Linux Performance Tools and How to Use Them

This article compiles a set of frequently used Linux performance utilities, originally sourced from Netflix's "Linux Performance Analysis in 60,000 Milliseconds" blog, and adds the author's own explanations.

1. uptime

$ uptime

The command shows overall system load; the three numbers after load average represent the average number of runnable processes over the last 1, 5, and 15 minutes.

2. dmesg | tail

$ dmesg | tail

Displays the tail of the kernel ring buffer, useful for spotting recent errors such as out‑of‑memory kills or dropped TCP packets.

3. vmstat 1

$ vmstat 1

Shows per‑second statistics for processes, memory, swap, I/O, and CPU. The first line is an average since boot; subsequent lines are sampled at the specified interval (1 s here). Important columns include:

r : runnable or waiting processes (CPU saturation if > CPU cores)

free : idle memory

si/so : swap in/out (non‑zero indicates memory pressure)

us, sy, id, wa : user, system, idle, and I/O wait CPU percentages (high wa suggests I/O bottlenecks)

4. mpstat -P ALL 1

$ mpstat -P ALL 1

Prints per‑CPU utilization each second, helping to verify whether load is evenly distributed across cores.

5. pidstat 1

$ pidstat 1

Shows CPU usage per process, similar to top but without screen clearing. Columns of interest:

%CPU : CPU percentage (values >100% indicate multi‑core usage)

PID and Command : identify offending processes.

6. iostat -xz 1

$ iostat -xz 1

Provides extended I/O statistics. Key columns:

r/s, w/s : reads/writes per second

await : average I/O wait time (high values signal saturation)

avgqu-sz : average queue size (values >1 may indicate overload)

%util : device utilization percentage (over 60% often means a bottleneck).

7. free -m

$ free -m

Shows memory usage. Important fields:

buffers : block device cache

cached : page cache (low values together with low free memory can indicate high I/O pressure).

8. sar -n DEV 1

$ sar -n DEV 1

Reports per‑second network statistics. Columns to watch:

IFACE : interface name

rxpck/s, txpck/s : packets received/transmitted per second

rxkB/s, txkB/s : kilobytes received/transmitted per second

await : average I/O wait for network devices.

9. sar -n TCP,ETCP 1

$ sar -n TCP,ETCP 1

Shows TCP connection activity. Useful columns:

active/s : locally‑initiated connections per second

passive/s : remotely‑initiated connections per second

retrans/s : TCP retransmissions per second.

10. top

$ top

A comprehensive, real‑time view of system load, process list, CPU, memory, and swap usage. Highlights include overall load average, per‑process CPU and memory percentages, and state counts (running, sleeping, zombie, etc.).

Summary

The accompanying diagrams illustrate each command’s primary purpose: vmstat for overall system health, mpstat for CPU distribution, pidstat for per‑process CPU usage, iostat for I/O performance, free for memory status, and sar for network monitoring.