Performance testing of OpenStack
Python Shell
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
ansible
conf
graphing
lib
log
rally
results
.gitignore
.gitreview
LICENSE
README.md
browbeat-complete.yaml
browbeat-config.yaml
browbeat.py
requirements.txt
setup.cfg

README.md

Note

This project has been migrated to the OpenStack namespace on Browbeat becoming an OpenStack Big Tent Project. All current development takes places at https://github.com/openstack/browbeat, which is the repostory you would want to clone. This repostory is no longer maintained. Table of Contents

Browbeat

This started as a project to help determine the number of database connections a given OpenStack deployment uses via stress tests. It has since grown into a set of Ansible playbooks to help check deployments for known issues, install tools, run performance stress workloads and change parameters of the overcloud.

Before running browbeat

  • Execute the ansible/gen_hostfile.sh script (builds the ssh config)
  • Configure browbeat-config.yaml to match your tests

How to run Browbeat Stress Tests?

On the Undercloud host, as the Stack user jump into the browbeat venv and you simply run:

(browbeat-venv)[stack@ospd browbeat]$ ./browbeat.py --help

However, the playbook required to install browbeat and its dependencies(Rally, Shaker, Perfkit) needs to be run before this. DEtailed install and run instructions are presented in a section below.

What is necessary?

  • TripleO
    • Why? We use passwordless ssh to reach each controller instance and compute instance.
  • 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). Other monitoring and stress test tools are installed by the respective playbooks when run.

If you prefer to not use Ansible, the older versions (no longer maintained) of the browbeat.sh can be found in a older commit.

Detailed Install, Check and Run

Installing Browbeat and running the Overcloud checks can be performed either from your local machine or from the undercloud. The local machine install/check assumes you have ansible installed already.

Install Browbeat from your local machine

From your local machine

$ ssh-copy-id stack@<undercloud-ip>
$ git clone https://github.com/jtaleric/browbeat.git
$ cd browbeat/ansible
$ ./gen_hostfile.sh <undercloud-ip> ~/.ssh/config
$ vi install/group_vars/all.yml # Make sure to edit the dns_server to the correct ip address
$ ansible-playbook -i hosts install/browbeat.yml
$ vi install/group_vars/all.yml # Edit browbeat network settings
$ ansible-playbook -i hosts install/browbeat_network.yml
$ ansible-playbook -i hosts install/shaker_build.yml

(Optional) Install collectd

$ ansible-playbook -i hosts install/collectd-openstack.yml

(Optional) Install collectd->graphite dashboards

$ ansible-playbook -i hosts install/dashboards-openstack.yml

(Optional) Install connmon

$ ansible-playbook -i hosts install/connmon.yml

Run Overcloud checks

$ ansible-playbook -i hosts check/site.yml

Your Overcloud check output is located in check/bug_report.log

NOTE: It is strongly advised to not run the ansible playbooks in a venv.

Run performance stress tests through browbeat on the undercloud:

$ ssh undercloud-root
[root@ospd ~]# su - stack
[stack@ospd ~]$ screen -S browbeat
[stack@ospd ~]$ . browbeat-venv/bin/activate
(browbeat-venv)[stack@ospd ~]$ cd browbeat/
(browbeat-venv)[stack@ospd browbeat]$ vi browbeat-config.yaml # Edit browbeat-config.yaml to control how many stress tests are run.
(browbeat-venv)[stack@ospd browbeat]$ ./browbeat.py <workload> #perfkit, rally, shaker or "all"

Install Browbeat directly on undercloud

From your undercloud

$ ssh undercloud-root
[root@ospd ~]# su - stack
[stack@ospd ~]$ git clone https://github.com/jtaleric/browbeat.git
[stack@ospd ~]$ cd browbeat/ansible
[stack@ospd ansible]$ ./gen_hostfile.sh localhost ~/.ssh/config
[stack@ospd ansible]$ sudo easy_install pip
[stack@ospd ansible]$ sudo pip install ansible
[stack@ospd ansible]$ vi install/group_vars/all.yml # Make sure to edit the dns_server to the correct ip address
[stack@ospd ansible]$ ansible-playbook -i hosts install/browbeat.yml
[stack@ospd ansible]$ vi install/group_vars/all.yml # Edit browbeat network settings
[stack@ospd ansible]$ ansible-playbook -i hosts install/browbeat_network.yml
[stack@ospd ansible]$ ansible-playbook -i hosts install/shaker_build.yml

(Optional) Install collectd

[stack@ospd ansible]$ ansible-playbook -i hosts install/collectd-openstack.yml

(Optional) Install collectd->graphite dashboards

[stack@ospd ansible]$ ansible-playbook -i hosts install/dashboards-openstack.yml

(Optional) Install connmon

[stack@ospd ansible]$ ansible-playbook -i hosts install/connmon.yml

Run Overcloud checks

[stack@ospd ansible]$ ansible-playbook -i hosts check/site.yml

Your Overcloud check output is located in check/bug_report.log

Run performance stress tests through browbeat

[stack@ospd ansible]$ . ../../browbeat-venv/bin/activate
(browbeat-venv)[stack@ospd ansible]$ cd ..
(browbeat-venv)[stack@ospd browbeat]$ vi browbeat-config.yaml # Edit browbeat.cfg to control how many stress tests are run.
(browbeat-venv)[stack@ospd browbeat]$ ./browbeat.py <workload> #perfkit, rally, shaker or "all"

Running PerfKitBenchmarker

Work is on-going to utilize PerfKitBenchmarker as a workload provider to browbeat. Many benchmarks work out of the box with browbeat. You must ensure that your network is setup correctly to run those benchmarks and you will need to configure the settings in ansible/install/group_vars/all.yml for browbeat public/private networks. Currently tested benchmarks include: aerospike, bonnie++, cluster_boot, copy_throughput(cp,dd,scp), fio, iperf, mesh_network, mongodb_ycsb, netperf, object_storage_service, ping, scimark2, and sysbench_oltp.

To run browbeat's PerfKit Benchmarks, you can start by viewing the tested benchmark's configuration in conf/browbeat-perfkit-complete.yaml. You must add them to your specific browbeat config yaml file or enable/disable the benchmarks you wish to run in the default config file (browbeat-config.yaml). There are many flags exposed in the configuration files to tune how those benchmarks run. Additional flags are exposed in the soruce code of PerfKitBenchmarker available: https://github.com/GoogleCloudPlatform/PerfKitBenchmarker

Example running only PerfKitBenchmarker benchmarks with browbeat from browbeat-config.yaml:

(browbeat-venv)[stack@ospd browbeat]$ ./browbeat.py  perfkit -s browbeat-config.yaml

Contributing

Contributions are most welcome! Pull requests need to be submitted using the gerrit code review system. Firstly, you need to login to GerritHub using your GitHub credentials and need to authorize GerritHub to access your account. Once you are logged in click you user name in the top-right corner, go to 'Settings' and under 'SSH Public Keys' you need to paste your public key. You can view your public key using:

$ cat ~/.ssh/id_\{r or d\}sa.pub

Set your username and email for git and gerrithub:

$ git config --global user.email "example@example.com"
$ git config --global user.name "example"
$ git config --global --add gitreview.username "example"

Next, Clone the github repository:

$ git clone https://github.com/jtaleric/browbeat.git

You need to have git-review in order to be able to submit patches using the gerrit code review system. You can install it using:

$ yum install git-review

To set up your cloned repository to work with gerrit:

[user@laptop browbeat]$ git review -s

Make your changes and then commit them. Use:

[user@laptop browbeat]$ git review

If you want to edit an already submitted patch, follow the below series of steps:

Firstly, go to the browbeat directory. Then,

git review -d Change-Id

Change-Id is the change id number as seen on gerrithub.io.

The above command downloads your patch onto a seperate branch. You might need to rebase your local branch with remoste master before running the abovecommand to avoid merge conflicts when you resubmit your edited patch. To avoid this, Go back to a "safe" commit using

$git reset --hard commit-number

Then,

$ git fetch origin
$ git rebase origin/master

Make the changes on the branch that was setup by using the git review -d (the name of the branch is along the lines of review/username/branch_name/patchsetnumber).

Add the files to git and commit your changes using,

$ git commit --amend

You can edit your commit message as well in the prompt shown upon executing above command.

Finally, push the patch for review using,

$ git review