x/browbeat
Code Issues Proposed changes
Performance monitoring and testing of OpenStack
140 Commits 1 Branch 0 Tags 23 MiB
Jinja 49.2%
Python 46.7%
Shell 3.7%
Dockerfile 0.4%
0201329526
Go to file
Joe 0201329526 Fix for haproxy
2015-12-07 21:05:55 -05:00
ansible Fix for haproxy 2015-12-07 21:05:55 -05:00
graphing Allow matplotlib to run without a screen. 2015-11-14 09:32:19 -05:00
keystone Fixed times in keystone-cc, adjusted concurrency of keystone tests, pad concurrency in task output log/html file. 2015-09-28 08:14:38 -04:00
neutron Adding Readme to directories 2015-09-23 10:01:59 -04:00
nova One more update 2015-09-24 15:49:43 -04:00
pbench Move install script to pbench folder. 2015-10-23 11:23:12 -04:00
rally-plugins Update README.md 2015-11-25 18:36:18 -05:00
browbeat-run-rally.sh Adding pbench to browbeat 2015-10-06 14:25:46 -04:00
browbeat.sh Added ansible clean logs playbook. 2015-11-25 16:18:44 -05:00
README.md Update README.md 2015-11-25 18:30:50 -05:00

README.md

Browbeat

Scripts to help determine number of workers a given OpenStack service needs.

How to run Browbeat?

On the Red Hat OpenStack Director host, as the Stack user jump into a venv w/ Rally and you simply run:

./browbeat.sh

What is necessary?

  • Red Hat OpenStack Director
    • Why? We use the keyless ssh to reach each controller node to maniuplate and check # of workers
  • OpenStack Rally
    • Why? We are using Rally to stress the control plane of the env.
  • Ansible
    • Why? We started with using bash to make changes to the Overcloud, creating complex sed/awks that we get for free with Ansible (for the most part). If you prefer to not use Ansible, the older versions (no longer maintained) of the browbeat.sh can be found here.