vmstat — Virtual Memory and CPU Statistics

vmstat reports a compact set of system activity metrics: runnable processes, memory, swap, I/O, interrupts, context switches, and CPU usage. Use it when WordPress feels slow and you suspect swapping, I/O wait, or CPU saturation. Unlike free, which is a snapshot, vmstat is most useful when sampled repeatedly (for example every second).

Quick Summary

Run vmstat 1 5 and watch si/so (swap in/out) and wa (I/O wait). Sustained swap activity or high I/O wait often correlates with slow page loads.

Reading the Columns (Mental Model)

vmstat-column-groups.txt

procs  -> r (run queue), b (blocked)
memory -> swpd, free, buff, cache
swap   -> si (swap in), so (swap out)
io     -> bi (blocks in), bo (blocks out)
system -> in (interrupts), cs (context switches)
cpu    -> us, sy, id, wa, st

Prerequisites

vmstat is provided by procps/procps-ng and is installed by default on most Linux servers.

Verify it exists:

verify-vmstat-installed.sh

which vmstat
vmstat -V

Core Syntax

vmstat-syntax.sh

vmstat [options] [delay [count]]

• delay is the sampling interval in seconds.
• count is the number of samples.

Key Options

Option	What it does	Example	WordPress / VPS use case
delay count	Sample repeatedly	vmstat 1 5	Observe real-time pressure
-w	Wide output	vmstat -w 1 5	Easier to read on modern systems
-t	Add timestamp	vmstat -t 1 5	Correlate with logs
-y	Skip the first report line	vmstat -y 1 5	Avoid the "since boot" first sample
-s	Summary statistics	vmstat -s	Baseline counters
-d	Disk statistics	vmstat -d	Quick disk activity counters
-p <dev>	Partition statistics	vmstat -p sda1	Target a specific partition
-a	Active/inactive memory	vmstat -a 1 5	Memory behavior under load

Examples (Commands + Expected Output)

Output varies

Column meanings stay consistent, but exact values depend on workload.

Sample every second (5 samples)

vmstat-sample-1s-5.sh

vmstat 1 5

Expected output:

example-output-vmstat-1-5.txt

procs -----------memory---------- ---swap-- -----io---- -system-- ------cpu-----
 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st
 1  0  20000 100000  60000 900000    0    0     1     2  120  300  5  2 90  3  0
 0  0  20000  90000  60000 910000    0    0     0     1  110  280  4  2 92  2  0

Use case: Spot swap (si/so) and I/O wait (wa) during slow requests.

Wide output

vmstat-wide-output.sh

vmstat -w 1 3

Expected output:

example-output-vmstat-wide.txt

procs -----------------------memory---------------------- ---swap-- -----io---- -system-- ----------cpu----------
 r  b       swpd       free       buff      cache   si   so    bi    bo   in   cs  us  sy  id  wa  st
 0  0      20000      90000      60000    910000    0    0     0     1  110  280   4   2  92   2   0

Use case: Better readability when columns would otherwise wrap.

Add timestamps

vmstat-with-timestamps.sh

vmstat -t 1 3

Expected output:

example-output-vmstat-timestamp.txt

... (columns omitted)
 0  0  20000  90000  60000 910000  0  0  0  1 110 280  4 2 92 2 0  2026-03-01 10:05:01

Use case: Correlate spikes with application logs.

Skip the first "since boot" report

vmstat-skip-first-line.sh

vmstat -y 1 3

Expected output:

example-output-vmstat-no-first.txt

... (only sampled lines, not the initial summary)

Use case: Focus on current behavior instead of boot-averaged counters.

Summary statistics

vmstat-summary-stats.sh

vmstat -s

Expected output:

example-output-vmstat-s.txt

     2048000 K total memory
      900000 K used memory
      100000 K free memory
       20000 K swap used

Use case: Baseline counters and totals.

Disk statistics

vmstat-disk-stats.sh

vmstat -d

Expected output:

example-output-vmstat-d.txt

disk- ------------reads------------ ------------writes----------- -----IO------
       total merged sectors      ms  total merged sectors      ms    cur    sec
sda   10000    200  800000   12000   5000    100  200000    8000      0     10

Use case: Quick disk activity counters (pair with iostat for deeper detail).

WordPress VPS Use Cases

Symptom	What to watch	Command	Interpretation
Site slow under load	wa and r	vmstat 1 10	High wa suggests disk bottleneck; high r suggests CPU saturation
Uploads failing	si/so and free	vmstat 1 10	Swap activity indicates memory pressure
DB-heavy queries slow	wa + bo	vmstat 1 10	Disk write pressure can slow MySQL

Troubleshooting

Observation	Likely cause	Next step
si/so stays > 0	Active swapping	Check memory with free -h; identify top memory processes
wa is consistently high	Disk I/O wait	Use iostat -xz 1 5 and review storage
r is high vs CPU cores	CPU contention	Check top/htop and per-CPU with mpstat

Best Practices

• Use vmstat as a time series (vmstat 1 10), not a single sample.
• Skip the first line with -y when you're focused on current behavior.
• Pair with free, top, and iostat to confirm the root cause.

Cheat Sheet

vmstat-cheat-sheet.sh

vmstat 1 5
vmstat -w 1 5
vmstat -t 1 5
vmstat -y 1 5
vmstat -s
vmstat -d
vmstat -p sda1
vmstat -a 1 5

What's Next

• free — Memory and Swap Usage
• top — Real-Time Process Monitoring
• iostat — Disk I/O Statistics
• sar — Historical Performance Statistics