Master Linux Performance: 5 Essential Monitoring Commands Explained

5 Linux Performance Monitoring Commands

1. vmstat – reports virtual memory statistics

vmstat prints detailed reports on memory, swap, I/O and CPU activity, including used/available memory, swap in/out, disk block reads/writes and CPU idle time.

Example: vmstat 5 Useful vmstat options:

-a – show active and inactive memory

-s – display event counters and memory statistics

-S – output in KB instead of blocks

5 – refresh output every 5 seconds

Each column meaning:

procs – process statistics

r: number of processes running or runnable

b: number of processes blocked

memory – memory statistics

swpd: used swap space (KB)

free: free memory (KB)

buff: memory used as buffers (KB)

cache: memory used as cache (KB)

swap – swap space statistics

si: KB/s swapped in from disk

so: KB/s swapped out to disk

io – I/O statistics

bi: blocks received per second

bo: blocks sent per second

system – system statistics

in: interrupts per second

cs: context switches per second

cpu – CPU statistics

us: % CPU time spent in user space

sy: % CPU time spent in kernel space

id: % CPU idle time

wa: % CPU time waiting for I/O

st: % stolen time in virtualized environments

2. iostat – reports CPU and I/O statistics

iostat monitors and displays CPU utilization and disk I/O metrics, including CPU load, IOPS, and read/write throughput.

Common iostat options:

c – display CPU usage information

t – print timestamp for each report

x – show extended statistics such as service time and wait counts

d – display detailed statistics for each disk/partition

p – display statistics for a specific disk device

Example (display detailed I/O for sda every 5 seconds): iostat -d -p sda 5 Sample output includes an avg‑cpu line (CPU usage percentages) and a Device line (tps, kB_read/s, kB_wrtn/s, etc.).

3. free – shows available and used memory

free displays total, used, and free amounts of physical and swap memory, providing an overview of memory availability.

Example: free -h Useful free options:

b – display output in bytes

k – display output in KB

m – display output in MB

h – display output in human‑readable format (GB, MB, etc.)

4. df – reports filesystem disk space usage

df shows filesystem name, total/used/available space and usage percentage.

Common df options:

-h – human‑readable format (K, M, G)

-T – display filesystem type

-i – show inode usage instead of space

-a – include all filesystems

-x type – exclude filesystems of the given type

-hT – combine -h and -T

--total – show total space usage in the last line

--help – display help

Example: df -h Columns meaning:

Filesystem – name of the filesystem

Type – filesystem type

1K-blocks – total number of 1‑KB blocks

Used – used blocks

Available – available blocks

Use% – percentage of space used

Mounted on – mount point

5. sar – collects and reports system activity

sar gathers and records CPU, memory, I/O, network and other system activity over time, enabling performance problem analysis.

Syntax: sar [options] [interval [count]] Common sar options:

-u – CPU utilization

-r – memory usage

-b – disk I/O

-n DEV – network interface statistics (DEV is interface name)

-q – run queue and load average

-A – all available statistics

-s HH:MM:SS – start time

-e HH:MM:SS – end time

-f FILE – specify sar data file to read

-o OUTPUT_FILE – write statistics to a file

Example (sample CPU usage every 5 seconds, 60 times):

sar -u 5 60