dev-lxc builds and manages clusters of LXC containers and includes the ability to install and configure Chef products.
Cluster management includes the ability to manage snapshots of the containers which makes dev-lxc well suited as a tool for support related work, customized cluster builds for demo purposes, as well as general experimentation and exploration.
- LXC Containers - Resource efficient servers with fast start/stop times and standard init
- Btrfs - Efficient, persistent storage backend provides fast, lightweight container snapshots
- Dnsmasq - DHCP networking and DNS resolution
- Base Containers - Containers that are built to resemble a traditional server
- ruby-lxc - Ruby bindings for liblxc
- YAML - Simple, flexible definition of clusters
- Build process closely follows online installation documentation
- Snapshots - Snapshots are created during the cluster's build process which makes rebuilding a cluster very fast.
- mixlib-install library - Automatically manages a cache of Chef products
Its containers, standard init, networking and build process are designed to be similar to what you would build if you follow the online installation documentation so the end result is a cluster that is relatively similar to a more traditionally built cluster.
The Btrfs backed snapshots provide a quick clean slate which is helpful especially for experimenting and troubleshooting. Or it can be used to build a customized cluster for demo purposes and be able to bring it up quickly and reliably.
If you aren't familiar with using containers you might be interested in this introduction.
Additional dev-lxc related documentation can be found in the docs folder in this repository.
The dev-lxc tool is used in a system that has been configured by the dev-lxc-platform cookbook.
The easiest way to build a dev-lxc-platform system is to download the dev-lxc-platform repository and use Test Kitchen to build an AWS EC2 instance or a VirtualBox Vagrant instance.
Follow the instructions in the dev-lxc-platform README to build a dev-lxc-platform instance. It takes about 7 minutes to build the dev-lxc-platform instance.
Login to the dev-lxc-platform instance and switch to the root user to use the dev-lxc tool.
cd dev-lxc-platform
kitchen login <ec2 or vagrant>
sudo -i
When you are logged in as the root user you should automatically enter a byobu session.
See byobu keybindings to learn some byobu basics.
Run the following command as the instance's root user if you ever need to upgrade the dev-lxc gem inside the dev-lxc-platform instance.
cd dev-lxc-platform
kitchen login <ec2 or vagrant>
sudo -i
chef gem update dev-lxc
dl is the dev-lxc command line tool.
dl subcommands and some options can be auto-completed by pressing the Tab key.
You only have to type enough of a dl subcommand to make it unique.
For example, the following commands are equivalent:
dl help
dl he
dl help
dl help <subcommand>
Cluster build time: 19 minutes
Create an Ubuntu 14.04 base container for the cluster's containers.
dl create b-ubuntu-1404
Create a directory to hold the dev-lxc.yml file.
mkdir -p /root/clusters/automate
The following command creates a dev-lxc.yml file that defines a standalone Chef Server, Supermarket server, Compliance server, Chef Automate server a Job Dispatch Runner and an infrastructure node and removes the "reporting" product from the generated configuration.
dl init --chef --compliance --supermarket --automate --runners --nodes --product-versions reporting:none > /root/clusters/automate/dev-lxc.yml
Copy your delivery.license file to the /root/clusters directory.
Run the cluster-view command to create a Byobu session specifically for this cluster.
cluster-view /root/clusters/automate
The session's first window is named "cluster".
The left pane is useful for running dev-lxc commands.
The right pane updates every 0.5 seconds with the cluster's status provided by dl status.
The session's second window is named "shell". It opens in the same directory as the
cluster's dev-lxc.yml file and is useful for attaching to a server to perform system administration tasks.
See the usage docs for more information about how to close/kill Byobu sessions.
dl up
At this point all of the cluster's servers should be running.
Since the cluster has a Chef Server and an infrastructure node dev-lxc made sure it configured the node's chef-client for the Chef Server so it is easy to converge the node.
You can use the attach subcommand to login to a server as the root user.
For example, the following commands should attach to node-1.lxc, start a chef-client run and exit the node.
dl attach node
chef-client
exit
Since the cluster has a Chef Server you can use the chef-repo subcommand to create a .chef directory in the host instance that contains a knife.rb and all of the keys for the users and org validator clients that are defined in dev-lxc.yml. This makes it very easy to use tools such as knife or berkshelf.
dl chef-repo
# set `username` to `mary-admin` and `orgname` to `demo` in `.chef/knife.rb`
knife client list
Since the cluster has a Chef Automate server you can use the print-automate-credentials subcommand to see the login credentials.
dl print
If you enabled local port forwarding for port 8888 in your workstation's SSH config file and configured your web browser to use 127.0.0.1:8888 for HTTP and HTTPS proxies as described in the dev-lxc-platform README.md then you should be able to browse from your workstation to any server that has a web interface using its FQDN.
For example, browse to https://automate.lxc and login with the credentials provided by dl print-automate-credentials.
Let's stop the servers before restoring and creating snapshots.
dl halt
You can restore a specific snapshot by name if you desire.
For example, the following command restores the Chef Automate server to the state right after its package was installed but before it was configured.
dl snapshot automate -r snap0
You can restore the most recent snapshot of all the servers.
dl snapshot -r
You can create snapshots with or without a comment.
dl snapshot -c 'Demo snapshot'
You can destroy snapshots.
dl snapshot -d snap2
Generally speaking, a cluster can be reused for a long time especially since snapshots easily allow you to restore the cluster to its initial build state. However, if you really want to destroy the servers and their snapshots you can use the destroy subcommand.
dl destroy
See the files in example-clusters.
For more in-depth documentation please see the pages in the docs folder.
- Fork it
- Create your feature branch (
git checkout -b my-new-feature) - Commit your changes (
git commit -am 'Add some feature') - Push to the branch (
git push origin my-new-feature) - Create new Pull Request