Docs
Command, Tool
vmstat

vmstat

This document summarizes the usage of vmstat, which displays memory statistics.

1. vmstat

1.1. vmstat 

$ vmstat
procs -----------memory---------- ---swap-- -----io---- -system-- ------cpu-----
 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st
 1  1    268 782960 849224 3488324    0    0    14  1733   77   78  6  2 68 24  0
[Shell 1] vmstat

Displays current memory statistics. [Shell 1] shows the output of vmstat displaying memory statistics. Each column has the following meaning:

  • procs
    • r : Number of Runnable Processes (Running state / waiting in Run Queue)
    • b : Number of Processes in Uninterruptible Sleep state
  • memory
    • swpd : Virtual memory capacity provided by Swap in bytes
    • free : Unused memory capacity in bytes
    • buff : Memory capacity used as Kernel Buffer in bytes
    • cache : Memory capacity used as Kernel Cache in bytes
  • swap :
    • si : Memory capacity swapped in from Swap Disk to Memory per second in bytes
    • so : Memory capacity swapped out from Memory to Swap Disk per second in bytes
  • io
    • bi : Number of blocks received from Block Device per second
    • bo : Number of blocks sent to Block Device per second
  • system
    • in : Number of interrupts received per second, including clock
    • cs : Number of Context Switches occurred per second
  • cpu
    • us : CPU usage rate for running application code
    • sy : CPU usage rate for running kernel code
    • id : CPU idle rate excluding I/O Wait
    • wa : CPU idle rate due to I/O Wait
    • st : CPU usage rate stolen by the hypervisor or other virtual machines when the kernel runs inside a virtual machine controlled by a hypervisor

1.2. vmstat [Interval] [Count]

Outputs system-wide information [Count] times at [Interval] intervals.

1.3. vmstat -d 

$ vmstat -d
disk- ------------reads------------ ------------writes----------- -----IO------
       total merged sectors      ms  total merged sectors      ms    cur    sec
loop0   8669      0   21324  109832      0      0       0       0      0      4
loop1     53      0    2152    1244      0      0       0       0      0      0
loop2  11152      0   25952     712      0      0       0       0      0      0
loop3      2      0      10       0      0      0       0       0      0      0
loop4      0      0       0       0      0      0       0       0      0      0
loop5      0      0       0       0      0      0       0       0      0      0
loop6      0      0       0       0      0      0       0       0      0      0
loop7      0      0       0       0      0      0       0       0      0      0
sr0        0      0       0       0      0      0       0       0      0      0
fd0       25      0     200    1164      0      0       0       0      0      1
sda   613771  66065 10319473 12671488 4981677 7462940 1628317680 348232712      0  94326
sdb      959      0   53207   20628      0      0       0       0      0     15
[Shell 2] vmstat -d

Displays disk statistics. [Shell 2] shows the output of vmstat -d displaying disk statistics. Each column has the following meaning:

  • reads
    • total : Total number of successful read operations
    • merged : Total number of read operations merged by I/O Scheduler
    • sectors : Total number of sectors successfully read
    • ms : Total time taken for reads
  • writes
    • total : Total number of successful write operations
    • merged : Total number of write operations merged by I/O Scheduler
    • sectors : Total number of sectors successfully written
    • ms : Total time taken for writes
  • IO
    • cur : Number of I/O operations currently being processed
    • sec : Total time taken for I/O processing in seconds

1.4. vmstat -p [Partition] 

# vmstat -p /dev/sda2
sda2          reads   read sectors  writes    requested writes
              612711   10283732    4783736 1628860584
[Shell 3] vmstat -p [Partition]

[Shell 3] shows the output of vmstat -p displaying partition statistics. Each column has the following meaning:

  • reads : Number of read requests issued for the partition
  • read sectors : Number of sectors read for the partition
  • writes : Number of write requests issued for the partition
  • requested writes : Number of sectors written for the partition

1.5. vmstat -m 

$ vmstat -m
Cache                       Num  Total   Size  Pages
SCTPv6                       22     22   1472     22
SCTP                          0      0   1344     12
au_finfo                      0      0    192     21
au_icntnr                     0      0    768     21
au_dinfo                      0      0    128     32
ovl_inode                 14122  14122    688     23
sw_flow                       0      0   1648     19
nf_conntrack                 94    168    320     12
ext4_groupinfo_4k          1624   1624    144     28
btrfs_delayed_ref_head        0      0    152     26
btrfs_delayed_node            0      0    296     13
btrfs_ordered_extent          0      0    416     19
btrfs_extent_map              0      0    144     28
btrfs_extent_buffer           0      0    280     14
btrfs_path                    0      0    112     36
btrfs_trans_handle            0      0    120     34
btrfs_inode                   0      0   1144     14
[Shell 4] vmstat -m

[Shell 4] shows the output of vmstat -m displaying slab statistics. Each column has the following meaning:

  • Cache : Cache name
  • Num : Number of currently active Slab Objects
  • Total : Total number of Slab Objects
  • Size : Size of Slab Object
  • Pages : Number of Pages (Slabs) with one or more active Slab Objects

2. References

top