Ps Utility in Linux (Procps), How to Check Which CPU Is Used

ps utility in linux (procps), how to check which CPU is used

which of multiple processors? it does not offer an option for that according to the manpage. but on my Debian stable system it accepts the undocumented -o cpu


after looking at the source, and the output of ps L, I believe your answer is either the cpuid or sgi_p output options, column IDs CPUID and P, respectively.

And 'cpu' should work according to this note in output.c, but it's currently tied to the 'nop' output pr_nop():

{"cpu", "CPU", pr_nop, sr_nop, 3, 0, BSD, AN|RIGHT}, /* FIXME ... HP-UX wants this as the CPU number for SMP? */

How to trace the list of PIDs running on a specific core?

TL;DR Dirty hacky solution.

DISCLAIMER: At some point stops working "column: line too long" :-/

Copy this to: core-pids.sh

#!/bin/bash

TARGET_CPU=0

touch lastPIDs
touch CPU_PIDs

while true; do
  ps ax -o cpuid,pid | tail -n +2 | sort | xargs -n 2 | grep -E "^$TARGET_CPU" | awk '{print $2}' > lastPIDs
  for i in {1..100}; do printf "#\n" >> lastPIDs; done
  cp CPU_PIDs aux
  paste lastPIDs aux > CPU_PIDs
  column -t CPU_PIDs > CPU_PIDs.humanfriendly.tsv
  sleep 1
done

Then

chmod +x core-pids.sh
./core-pids.sh

Then open CPU_PIDs.humanfriendly.tsv with your favorite editor, and ¡inspect!

The key is in the "ps -o cpuid,pid" bit, for more detailed info, please comment. :D

Explanation

Infinite loop with

  • ps -o cpuid,pid | tail -n +2 | sort | xargs -n 2 | grep -E "^$TARGET_CPU" | awk '{print $2}' > lastPIDs

    • ps ax -o cpuid,pid

      • Show pid's associated to CPU
    • tail -n +2

      • remove headers
    • sort

      • sort by cpuid
    • xargs -n 2

      • remove white spaces at begging
    • grep -E "^$TARGET_CPU"

      • filter by CPU id
    • awk '{print $2}'

      • get pid column
    • > lastPIDs

      • output to file those las pid's for the target CPU id
  • for i in {1..10}; do printf "#\n" >> lastPIDs; done

    • hack for pretty .tsv print with the "columns -t" command
  • cp CPU_PIDs aux

    • CPU_PIDs holds the whole timeline, we copy it to aux file to allow the next command to use it as input and output
  • paste lastPIDs aux > CPU_PIDs

    • Append lastPIDs columns to the whole timeline file CPU_PIDs
  • column -t CPU_PIDs > CPU_PIDs.humanfriendly.tsv

    • pretty print whole timeline CPU_PIDs file

Attribution

  • stackoverflow answer to: ps utility in linux (procps), how to check which CPU is used

    • by Mikel
  • stackoverflow answer to: Echo newline in Bash prints literal \n

    • by sth
  • stackoverflow answer to: shell variable in a grep regex

    • by David W.
  • superuser answer to: Aligning columns in output from a UNIX command

    • Janne Pikkarainen
  • nixCraft article: HowTo: Unix For Loop 1 to 100 Numbers

Bash script checking cpu usage of specific process

The problem is that bash can't handle decimals. You can just multiply them by 100 and work with plain integers instead:

#!/bin/bash
declare -i app_pid
declare -i app_cpu
declare -i cpu_limit
app_name="top"
cpu_limit="5000"
app_pid=`ps aux | grep $app_name | grep -v grep | awk {'print $2'}`
app_cpu=`ps aux | grep $app_name | grep -v grep | awk {'print $3*100'}`
if [[ $app_cpu -gt $cpu_limit ]]; then
     echo "crap"
else
     echo "we're good"
fi

Keep in mind that CPU percentage is a suboptimal measurement of application health. If you have two processes running infinite loops on a single core system, no other application of the same priority will ever go over 33%, even if they're trashing around.

how to redirect ps command in linux to look into other folder than /proc

the /proc path is hardcoded in the source tree of the /bin/ps binary file.
Thus, you need to recompile /bin/ps

Follow these steps to recompile /bin/ps inside a container that mounts the host /proc and use this new ps to display the process list of the docker host from the container:

In this example, I do not use /host/proc but /prod, to avoid modifying the path length to the mount point of the procfs filesystem (increasing the path length could trigger runtime errors in some situations). I also use a container based on OpenSUSE Leap 42.1, since you have not described the base image you are using.

1- on the docker host, first download the /bin/ps sources:

fenyo@myhost# mkdir /root/git
fenyo@myhost# cd /root/git
fenyo@myhost# git clone https://gitlab.com/procps-ng/procps.git
fenyo@myhost# cd procps

2- replace any occurence of "/proc by "/prod in **/*.c

3- run your docker container, mounting procfs on /prod in the container

fenyo@myhost# docker run -v /proc:/prod -v /root/git/procps:/root/git/procps -t -i --rm opensuse:42.1 bash

4- inside the docker container, recompile /bin/ps

bash-4.2# cd /root/git/procps
bash-4.2# zypper install -y gettext gettext-tools autoconf libtool pkg-config gcc make
bash-4.2# ./autogen.sh
bash-4.2# ./configure --without-ncurses
bash-4.2# make

Now, you can use /root/git/procps/ps/pscommand as a replacement for /bin/ps, in order to use /prod instead of /proc. With pscommand, you will get a process list from the host, not from the container:

bash-4.2# ps -auxgww
USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
root         1  0.0  0.0  20064  3076 ?        Ss   23:20   0:00 bash
root      5199  0.0  0.0  33340  2996 ?        R+   23:24   0:00 ps -auxgww

bash-4.2# /root/git/procps/ps/pscommand -auxgww
USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
root         1  0.0  0.0  33620  3956 ?        Ss   Jul03   1:13 /usr/lib/systemd/systemd --switched-root --system --deserialize 21
root         2  0.0  0.0      0     0 ?        S    Jul03   0:01 [kthreadd]
root         3  0.0  0.0      0     0 ?        R    Jul03   4:44 [ksoftirqd/0]
root         5  0.0  0.0      0     0 ?        S<   Jul03   0:01 [kworker/0:0H]
root         7  0.0  0.0      0     0 ?        S    Jul03  15:27 [rcu_preempt]
root         8  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcu_sched]
root         9  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcu_bh]
root        10  0.0  0.0      0     0 ?        S    Jul03  10:02 [rcuop/0]
root        11  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcuos/0]
root        12  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcuob/0]
root        13  0.0  0.0      0     0 ?        S    Jul03   0:06 [migration/0]
root        14  0.0  0.0      0     0 ?        S    Jul03   0:07 [watchdog/0]
root        15  0.0  0.0      0     0 ?        S    Jul03   0:06 [watchdog/1]
root        16  0.0  0.0      0     0 ?        S    Jul03   0:04 [migration/1]
root        17  0.0  0.0      0     0 ?        S    Jul03   1:07 [ksoftirqd/1]
root        19  0.0  0.0      0     0 ?        S<   Jul03   0:00 [kworker/1:0H]
root        20  0.0  0.0      0     0 ?        S    Jul03   2:42 [rcuop/1]
root        21  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcuos/1]
root        22  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcuob/1]
root        23  0.0  0.0      0     0 ?        S    Jul03   0:05 [watchdog/2]
root        24  0.0  0.0      0     0 ?        S    Jul03   0:06 [migration/2]
root        25  0.0  0.0      0     0 ?        S    Jul03   0:48 [ksoftirqd/2]
root        27  0.0  0.0      0     0 ?        S<   Jul03   0:00 [kworker/2:0H]
root        28  0.0  0.0      0     0 ?        S    Jul03   7:32 [rcuop/2]
root        29  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcuos/2]
root        30  0.0  0.0      0     0 ?        S    Jul03   0:00 [rcuob/2]
root        31  0.0  0.0      0     0 ?        S    Jul03   0:05 [watchdog/3]
root        32  0.0  0.0      0     0 ?        S    Jul03   0:05 [migration/3]
root        33  0.0  0.0      0     0 ?        S    Jul03   0:37 [ksoftirqd/3]
[...]

How to store ps command output in csv format?

You could try something like the following code I wrote:

ps -e -o %p, -o lstart -o ,%C, -o %mem -o ,%c > output.csv

Brief explanation:

The -o option can be used multiple times in a ps command to specify the format.
In order to control which separator is used we can use AIX format descriptors. We can specify our needed separators like, e.g. %p,. Since AIX format descriptors are not available for every piece of data, but only for some of the data (for example in our case there are no AIX format descriptors for %mem and for lstart), we plant %mem and lstart around the available AIX format descriptors to achieve the comma separation. For example this site provides information about the ps command for further readings.

output.csv example:

  PID,                 STARTED,%CPU,%MEM,COMMAND
    1,Mon Feb 25 00:00:01 2019, 0.0, 0.1,examplecommand1
    2,Mon Feb 25 00:00:01 2019, 0.0, 0.0,examplecommand2
                           (...)

Related Topics

In Linux, What Do All the Values in the "Top" Command Mean

How to Skip Saturday and Sunday in a Cron Expression

How to Open a "-" Dashed Filename Using Terminal

How to Determine the Current Ip from a Known MAC Address

How to Mount an Iso Inside a Docker Container

Mongodb Service Will Not Start After Initial Setup

How to Do Http-Request/Call with JSON Payload from Command-Line

How to Set Up Autocompletion for Git Commands

Qstat and Long Job Names

How to Open Sublime Text 2 Files from the Command Line in Linux to a Tab, Not a New Window

Does Epoll(), Do Its Job in O(1)

"Docker Images" Shows Image, "Docker Rmi" Says "No Such Image" or "Reference Does Not Exist"

Scp Command Syntax for Copying a Folder from Local MAChine to a Remote Server

How Does Epoll's Epollexclusive Mode Interact with Level-Triggering

Id_Rsa.Pub File Ssh Error: Invalid Format

Issue While Validating Bash Script

Why Using Dirname in Find Command Gives Dots for Each Match

How to Use Homebrew on Ubuntu


Leave a reply



Submit