vmstat

Display virtual memory statistics.

Description

The vmstat command stands for "Virtual Memory Statistics," but it can report overall system status regarding processes, memory, I/O, and more.

Syntax

vmstat (option) (parameter)

Options

-a: Display active memory pages;
-f: Display the total number of processes created since boot;
-m: Display slab information;
-n: Display the header only once;
-s: Display event counters and memory statistics in a table format;
-d: Report disk statistics;
-p: Display statistics for a specified hard disk partition;
-S: Specify the units for output information.

Parameters

• Interval: The time interval between refreshing status information;
• Count: The number of reports to display.

Examples

vmstat 3
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    320  42188 167332 1534368    0    0     4     7    1    0  0  0 99  0  0
 0  0    320  42188 167332 1534392    0    0     0     0 1002   39  0  0 100  0  0
 0  0    320  42188 167336 1534392    0    0     0    19 1002   44  0  0 100  0  0
 0  0    320  42188 167336 1534392    0    0     0     0 1002   41  0  0 100  0  0
 0  0    320  42188 167336 1534392    0    0     0     0 1002   41  0  0 100  0  0

Field Descriptions:

Procs (Processes)

• r: The number of processes in the run queue. This value can help determine if more CPUs are needed (if consistently > 1).
• b: The number of processes waiting for I/O.

Memory

• swpd: Amount of virtual memory used. If swpd is non-zero but SI and SO are consistently zero, system performance is generally unaffected.
• free: Amount of idle physical memory.
• buff: Amount of memory used as buffers.
• cache: Amount of memory used as cache. A high cache value indicates many files are cached; if frequently accessed files are cached, disk read I/O (bi) will be very low.

Swap

• si: Amount of memory paged in from swap (disk to memory) per second.
• so: Amount of memory paged out to swap (memory to disk) per second.

Note: When there is enough memory, both values are 0. If these values are consistently greater than 0, system performance may be affected as disk I/O and CPU resources are consumed. Do not worry if free memory is low as long as si and so remain low (mostly 0).

IO (Block size is typically 1KB in current Linux versions)

• bi: Blocks received from a block device (read) per second.
• bo: Blocks sent to a block device (write) per second.

Note: During random disk I/O, higher values (e.g., > 1024K) will lead to higher CPU I/O wait times.

System

• in: Number of interrupts per second, including the clock.
• cs: Number of context switches per second.

Note: Higher values for these two fields result in more CPU time consumed by the kernel.

CPU (Expressed as percentages)

• us: Percentage of time spent running non-kernel code (user time). High us values mean user processes are consuming a lot of CPU; if consistently over 50%, consider optimizing the program.
• sy: Percentage of time spent running kernel code (system time). High sy values indicate the kernel is consuming significant CPU resources, which should be investigated.
• wa: Percentage of time spent waiting for I/O. High wa values indicate serious I/O waiting, possibly due to heavy random disk access or disk bottlenecks.
• id: Percentage of time spent idle.