Essential Linux Commands for Quick System Performance Diagnosis

When you encounter performance problems on a Linux server, the first minute of investigation should focus on a set of core system metrics.

Uptime

The uptime command shows the system's average load over the past 1, 5, and 15 minutes, indicating how many tasks are waiting to run or are blocked on I/O.

dmesg | tail

Running dmesg | tail displays the latest kernel messages, helping you spot out‑of‑memory kills, hardware errors, or network issues such as SYN flooding.

vmstat 1

vmstat 1

reports virtual memory and CPU statistics every second. Key fields include r (runnable tasks), free (free memory in KB), si/so (swap activity), and CPU time breakdown ( us, sy, id, wa, st ).

mpstat -P ALL 1

mpstat -P ALL 1

prints per‑CPU utilization, allowing you to see whether a single core is saturated, which often points to a single‑threaded workload.

pidstat 1

pidstat 1

provides per‑process CPU usage at regular intervals, making it easy to identify processes that consume many CPU cores (e.g., a Java process showing >1500% CPU on a 32‑core box).

iostat -xz 1

iostat -xz 1

shows block‑device performance. Important metrics are r/s, w/s, rkB/s, wkB/s (I/O load), await (average request latency), avgqu‑sz (queue length), and %util (device utilization).

free -m

free -m

reports memory usage. Pay attention to the buffers and cached columns and the adjusted line -/+ buffers/cache, which gives a more accurate view of truly used memory.

sar -n DEV 1

sar -n DEV 1

monitors network interface throughput (rxkB/s, txkB/s) and utilization, helping you determine whether network bandwidth is a limiting factor.

sar -n TCP,ETCP 1

sar -n TCP,ETCP 1

reports TCP statistics such as active connections, passive connections, and retransmissions, useful for assessing connection‑level load and potential network issues.

top

The top command provides a real‑time view of processes, CPU usage, memory consumption, and load averages, allowing you to verify whether the metrics observed with the previous tools have changed.

These commands together form a quick, one‑minute checklist for diagnosing Linux server performance problems.