libra/doc/install/development.rst

Development environment

Libra is a system to provide LoadBalancing as a Service on top of various platforms. It is comprised of four components libra-api, libra-admin-api, libra-pool-mgm and libra-worker, supported by a few other open source components. For more information see /architecture/index.

Development environment

This guide will walk you through howto setup a development environment for Libra using:

• 1 Node for the API, Admin API and Pool mgm with MySQL
• n+ Nodes for workers that is going to run HAProxy
• Ubuntu 12.04 as our OS of choice

Common steps

1. Install general system dependencies

$ sudo apt-get install -qy python-virtualenv python-pip python-dev git gcc

2. Clone the repo from Stackforge at GitHub

$ git clone ssh://git@github.com/stackforge/libra
$ cd libra

3. Create needed directories

$ sudo mkdir -p /var/run/libra /var/log/libra
$ sudo chown ubuntu:ubuntu /var/run/libra /var/log/libra

Installing

double: install; libra

1. Do steps in ppa
2. Do steps in 'Common steps'
3. Install dependencies

$ sudo apt-get install -qy gearman-jobs-server mysql-server

4. Setup a VirtualEnvironment

Note

This is to not interfere with systemwide libraries.

$ virtualenv .venv
$ . .venv/bin/active

5. Install python-gearman

Note

This is a custom version with patches commited upstream but not release yet.

$ pip install https://launchpad.net/~libra-core/+archive/ppa/+files/gearman_2.0.2.git3.orig.tar.gz

6. Install dependencies using pip

$ pip install -r requirements.txt -r test-requirements.txt

7. Install python-keystoneclient

::

pip install python-keystoneclient

8. Install Libra in development mode

$ python setup.py develop

9. Copy the configuration file to /etc

$ sudo cp etc/sample_libra.cfg /etc/libra.cfg

10. Change running as libra to ubuntu

Note

This is to not have to add a new user.

$ sudo sed -r -i 's/^(group|user).*libra/\1 = ubuntu/' /etc/libra.cfg

11. Configure libra

$ sudo vi /etc/libra.cfg

Note

See configuration for how to proceed for various options.

You should at least configure the variables needed for your environment.

Setup database and gearman

1. Import the initial database

$ mysql < libra/common/api/lbaas.sql

2. Change the listening address of Gearman server

$ sudo vi /etc/default/gearman-job-server

3. Restart gearman

$ sudo service gearman-job-server restart

Bring up services

1. Start the Pool Manager

$ libra-pool-mgm -c /etc/libra.cfg

2. Start Admin API & API services

$ libra-admin-api -c /etc/libra.cfg
$ libra-api -c /etc/libra.cfg

Creating a Worker Image

Note

In this setup we'll be using OpenStack as the underlying provider for our Libra Worker nodes to run HAProxy on.

1. Boot a server using Nova

Note

You should at least open (for now at least) port 22 for ssh.

--nic argument is only needed if you have multiple networks. --security-groups is not needed at the time if you have 22 in default

$ nova boot --flavor <flavour id or name> --image <image id of ubuntu precise> --key-name default --nic net-id=<network id> --security-groups=<your security groups> worker

2. Create a floating ip

$ neutron floatingip-create <external network name>

3. Assign a floating ip to the instance

Note

You can view all the ports by issuing neutron port-list.

$ neutron floatingip-associate <floating ip id> <port id>

4. Login to the instance

$ ssh ubuntu@<ip>

5. Do steps in 'Common steps'
6. Install HAProxy

$ apt-get install -qy haproxy socat

7. Install python-gearman

Note

This is a custom version with patches commited upstream but not release yet.

sudo pip install https://launchpad.net/~libra-core/+archive/ppa/+files/gearman_2.0.2.git2.orig.tar.gz

8. Install dependencies using pip

$ sudo pip install -r requirements.txt -r test-requirements.txt

9. Install Libra in development mode

$ sudo python setup.py develop

10. Install a Upstart job

$ sudo wget https://raw.github.com/pcrews/lbaas-salt/master/lbaas-haproxy-base/libra_worker.conf -O /etc/init/libra_worker.conf

11. Make a snapshot of the worker image

$ nova image-create worker libra-worker

12. At the libra-poo-mgm node change the 'nova_image' setting to the value of your newly created snapshot

Note

To get the ID of the snapshot do glance image-show libra-worker | grep -w id | cut -d '|' -f3

$ sudo vi /etc/libra.cfg

13. Restart libra-pool-mgm

$ killall -9 libra_pool_mgm
$ libra_pool_mgm -c /etc/libra.cfg

Verifying that it works

If you have done all correctly you should be able to do something like the below command on the node that has the libra-pool-mgm

$ tail -f /var/log/libra/libra_mgm.log