How To Identify The Virtual Memory Status | vmstat Admin Monitoring Command

VMSTAT –Monitor Virtual Memory
vmstat is used to report the virtual Memory status.
vmstat report About the
1)Process,
2)Memory,
3)Swap
4)IO ,
5)System
6)CPU Performance.

These all the things are included in the Virtual Memory.
By through these monitoring we can identify & known How much virtual memory is there.
To Clear the Page Cache , dentries & inodes 
# echo 3 > /proc/sys/vm/drop_caches
 
Comparison between Cache and Buffer
Comparison between Paging and swapping Process Section
r Runnable Process
b Blocked Process
Memory Section (Ram, Buffer,Cache)
swpd>> swap Space
Free>> Available Ram
Buff>> Ram used for Buffer
Cache>> Ram used for cache
Swap Section
si (amount of)swapped from disk per second - shows page in
so swapped to disk per second - shows page out
IO Section
Bi (amount of)Block receive from disk
Bo Block sent to disk

System Section
In (no of) interrupt per second
Cs Context Switches Per second

CPU Section
us Time (spent) user mode (non kernel code)
sy Time kernel code
id idle time
wa waiting time


[root@ubuntu ~]# 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  0      0  10612 111764 202520    0    0   246    29 1040  240  5  7 85  2  0
Important While Monitoring
Memory>> free it will shows How My remaining idle memory space are there.
CPU >>id it will display how much CPU is busy
Si SO>> it is always display zero value kernel does not have physical memory – Paging out
Display Disk Table
vmstat -D option it will print the Disk table.
about the partiton, Disk,
merged reads, milli reading
merged writes, milli writing,
[root@ubuntu ~]# vmstat -D
20 disks
3 partitions
27096 total reads
25317 merged reads
964401 read sectors
206355 milli reading
4610 writes
10403 merged writes
119858 written sectors
364569 milli writing
0 inprogress IO
88 milli spent IO

Insert The Timestamp With Output
[root@ubuntu ~]# vmstat 1 3 | awk '{print strftime("%D %H:%M:%S",systime()) " "$0}'
05/20/10 06:33:30 procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu------
05/20/10 06:33:30 r b swpd free buff cache si so bi bo in cs us sy id wa st
05/20/10 06:33:30 2 0 0 8596 111852 203192 0 0 169 21 1036 230 4 6 88 1 0

Delay And The No Of Count
#vmstat 1 5
1 > delay - It will update there output Every 1 minutes
5 > five count of display

Related post