Master Linux Performance Monitoring: Essential Tools and Commands

This document consolidates Linux performance analysis tools, originally inspired by Brendan Gregg's work, and serves as a reference for system observability, benchmarking, and tuning.

Basic Monitoring Tools

vmstat : Shows virtual memory, processes, CPU, and I/O statistics. Usage: vmstat interval times.

iostat : Reports CPU and device I/O statistics. Use iostat -x for extended device stats.

dstat : Displays CPU, disk I/O, network, and paging in a colorful, detailed format. Example: dstat -cdlmnpsy.

iotop : Real‑time disk I/O monitoring per process, similar to top. Non‑interactive mode: iotop -bod interval.

pidstat : Monitors resource usage of specific processes (CPU, memory, I/O). Example: pidstat -d interval.

top / htop : Interactive process viewers showing load, CPU, memory, swap, and per‑process details.

mpstat : Multiprocessor CPU statistics, e.g., mpstat -P ALL interval times.

netstat : Network connection and routing information. Common options: netstat -npl, netstat -rn, netstat -i.

ps : Displays current process states. Typical usage: ps aux or ps -ef | grep pattern.

strace : Traces system calls and signals of a program, useful for debugging.

uptime : Shows system uptime and load averages.

lsof : Lists open files and the processes using them.

Advanced Tracing and Profiling

perf : Kernel‑level profiling tool for hotspots, cache misses, and function‑level performance.

eBPF tools (bcc, perf‑tools, ktap) : Modern tracing frameworks that leverage BPF for kernel and user‑space instrumentation.

Flame Graphs : Visual representation of stack samples generated by perf or eBPF.

Reference Images