PT - Monitoring PS command

 Monitoring Processes through PS command 

• ps -ef 

• PS command contains CPU cloumns

ps aux

• Top 10 process consuming CPU

ps aux | sort -nrk 3,3 | head -n 10

• Top10 process consuming CPU

 ps -e -o pcpu,pid,user,tty,args | sort -n -k 1 -r | head -n 10

PT - OS TOP Monitoring commands

 

PT - OS TOP commands Monitoring 

Top Command Header 

Top  | users | load average | Tasks | cpu  | Memory | swapping 

Field/Section	Meaning
load average	This line displays the load averages over the last 1, 5 and 15 minutes. The higher the load average, the more work the machine is doing. A high rise in the load average could indicate a serious performance problem.
Tasks	Number of processes that are running, sleeping, stopped, or zombie. In a healthy system, this number does not change a lot by time.
Cpu(s)	CPU% usage by user (%us), system (%sy), nice (%ni), idle (%id), I/O wait (%wa), Hard Interrupt (%hi), soft interrupt (%si), and iosteal (%st) which is used in VM environments. Note: in top output, %wa is the only I/O load indicator.
Mem	Total physical memory, used memory, and free memory.
Swap	Information about swapping.
PID	Process Id. Used to know the relationship between the processes and for killing processes
USER	The effective user name of the task’s owner.
PR	Process priority
NI	Process nice value
VIRT	Total virtual memory used by the process
RES	Non-swapped physical memory used
SHR	Shared memory used by the process
S	Process status
%CPU	Processes percent of CPU consumption since last screen refresh
%MEM	Percentage of memory consumption by the process
TIME+	Total CPU time, showing hundredths of seconds
COMMAND	Command used to start the process

Key	Impact
[Shift]+[O]	Sort by a selected field (try sorting by memory, swapped size, command name)
[Shift]+[P]	Sort processes as per CPU utilization
z	highlight running processes. Press again to disable it.
c	Display the absolute path of running processes
d	Change screen refresh interval (by default it is 3)
k	Kill a process by its PID
q	Quit

• Top command based on particular user 

top -u oracle

• Display first 20 line output.

top -bc | head -20 display first 20 lines 

• Display order by memory usage 

top -bc -a | head -n 20 

• Enter Z  highlighting current running process.

PT - OS performance Monitoring using VMSTAT

 

PT - OS performance Monitoring using  VMSTAT 

vmstat

process ---- Memory ---- swap ---- i/o ----- system ----  cpu 

Process : r | b

Memory : spwd | spwd | free | buffer | cache

Swap : si | so

I/O:  bi | bo

System : cs | us | sy 

CPU : id |wt | st 

Field

Meaning

r

number of processes waiting for run time

b	number of processes in uninterruptible sleep

spwd

Virtual memory swapped out to disk

spwd

Virtual memory swapped out to disk

free

Amount of free memory

si	Amount of memory swapped in from disk
so	Amount of memory swapped to disk
us	User CPU time
bi	Blocks received from a block        device (blocks/s)
bo	Blocks sent to a block device
sy	System CPU time
id	CPU idle time

• vmstat command output example 

• vmstat 2 ( output every 2 second)

• Example to apply load on OS by executing i/o 30 times using apply_io_stress.sh file.

cat > apply_io_stress.sh <<EOL

#!/bin/bash

counter=1

while [ \$counter -le 30 ]

do

dd if=/dev/zero of=/tmp/test1.img bs=500M count=1 oflag=dsync

((counter++))

done

rm -f /tmp/test1.img

EOL

• Provide execute permission below.

chmod +x apply_io_stress.sh

srv1:~ # sh apply_io_stress.sh

^Z

[1]+  Stopped                 sh apply_io_stress.sh

srv1:~ #

• We can see after executing sh file number of r b  process increases also I/O bi bo  reading blocks and write block increase.

^Z

[1]+  Stopped                 sh apply_io_stress.sh

• Once stopping apply_io_stress.sh file processes and I/O value came back to normal.

vmstat -d 2

• vmstat -d 2 ( Disk output at every 2 seconds) Examples below.