https://github.com/cscfi/ansible-role-stress

Run a stress test on a machine using Ansible
https://github.com/cscfi/ansible-role-stress

Last synced: 8 months ago
JSON representation

Run a stress test on a machine using Ansible

• Host: GitHub
• URL: https://github.com/cscfi/ansible-role-stress
• Owner: CSCfi
• License: mit
• Created: 2016-01-26T13:10:06.000Z (over 10 years ago)
• Default Branch: master
• Last Pushed: 2024-10-10T09:16:08.000Z (over 1 year ago)
• Last Synced: 2025-04-24T07:43:44.155Z (about 1 year ago)
• Language: Shell
• Size: 307 KB
• Stars: 3
• Watchers: 10
• Forks: 7
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
[![Build Status](https://travis-ci.org/CSCfi/ansible-role-stress.svg?branch=master)](https://travis-ci.org/CSCfi/ansible-role-stress)

# ansible-role-stress

Run a stress test on a machine using Ansible. This role uses

[stress-ng](http://kernel.ubuntu.com/~cking/stress-ng/) to stress the CPU,

memory and HDD disk(s), and [FIO](https://fio.readthedocs.io/en/latest/fio_doc.html)

to stress SSD(s) on a system. It installs stress-ng and fio and runs it wrapped on a

script with configurable parameters. Since running a stress test is something that is

generally done for a long period continuously, screen is used to contain

the script so that the Ansible run finishes in a timely manner and simply leaves

the stress test running in the background.

Requirements

------------

  * Only tested on CentOS 7. Should also work on Ubuntu

Role variables

--------------

  * stress_ng_duration: sets the timeout limit for stress-ng in seconds

  * Variables to set the number of each type of worker:

    * cpu_workers

    * vm_workers

    * hdd_workers

  * Variables to set how much memory and disk space to consume per worker:

    * bytes_per_hdd_worker

    * bytes_per_vm_worker

  * Variables to set how the SSD test (FIO) is done:

    * ssd_test: Test SSD disks using FIO (if there are SSD disks)

    * fio_test_duration: sets the timeout limit for fio test in seconds

    * ssd_block_size: Size of each block for FIO's SSD test

    * ssd_workers: Number of workers for FIO's SSD test

    * ssd_tests_folder: Folder to mount a new Logical Volume in the Volume Group vg_instances and where FIO would store the files for the test

    * ssd_volume_group: Volume Group name to test the SSDs. Check that this is contained in the SSDs

Notes

----

   * Up until version 2019-01-02 this role also supported CentOS 6. After that we no longer support that as we assume stress-ng is available in a yum repo configured on the server.