Tag Archives: mcollective

Role based Puppet/MCollective in EC2 – Part 3 Installing MCollective with SSL

This post will take you through adding MCollective orchestration software to the puppetmaster you just finished building. This howto includes configuring MCollective over SSL. This builds on the puppetmaster you built in Part 2. Ubuntu 14.04 running on EC2. On a side note, if you have the budget, I highly suggest supporting Puppet Labs and their efforts by subscribing to Puppet Enterprise. Of course if you did, you wouldn’t need these instructions 🙂

About MCollective.

MCollective is orchestration software by Puppet Labs. It allows you to connect to and run “jobs” simultaneously across groups of instances. These groups of instances can be manually defined or selected based on something like the Facter facts we defined in the user-data in Part 1 MCollective consists of 3 basic components:

  • Middleware: This is really the “server” portion. ActiveMQ messaging software runs here. I run this on my puppetmaster.
  • Agent: mcollectived runs on all the instances you want to control with mcollective.
  • Client: mco command. This is where you control the agents. I think of this part as the command console. I also install this on the puppetmaster.

Apache ActiveMQ is really the “server” portion of MCollective. You can use other messaging servers, but that’s way beyond the scope of what is covered here. It’s written in Java.

  • Listening Port: 61614
  • Start, Stop, Restart: service activemq [action]
  • Configuration: /etc/activemq/instances-enabled/mcollective/
  • Logs: Not setup by default. ActiveMQ Logging

mcollectived is the “agent” portion. It connects to the ActiveMQ server, then the server can pass messages back. It’s written in Ruby.

  • Start, Stop, Restart: service mcollective [action]
  • Configuration: /etc/mcollective
  • Logs: /var/log/mcollective.log

mco – the “client”
MCollective calls the places where you run the mco command the “client”. I think this is terribly confusing, it’s too close to “agent”. I generally install mco on my puppetmaster instances. Be careful who you give access to the mco command – it can be the equivalent of giving root access to all your instances. Mistakes here can affect all of your instances at once. It is possible to lock down what a user can do with this command, but that’s beyond the scope of this tutorial. This is also written in Ruby.

Installing MCollective.

I find the Puppet Labs documentation extremely confusing and out of date. But there is hope! I just about had it all figured out when I discovered that you can do the full installation of middleware/agent/client with SSL though the puppetlabs/mcollective Puppet module. Using the module almost makes the install easy.

Install puppetlabs/mcollective Puppet module.

Install the module with the puppet command. It will install a bunch of dependencies.

puppet module install puppetlabs/mcollective
Install my site_mcollectve Module.

I wrote a module that wraps the puppetlabs/mcollective module. Mostly the site_mcollective module gives you a place keep the certs and define your users so Puppet can install them.

Clone the Repo into `/etc/puppet/modules`.
I put the module up on GitHub – https://github.com/jgreat/site_mcollective

cd /etc/puppet/modules
git clone https://github.com/jgreat/site_mcollective.git

Setup Config File.
All the configuration is done in a yaml file. This will require Hiera to work. Create a folder to store yaml configs for Hiera and copy the site_mcollective.yaml to it.

mkdir /etc/puppet/yaml
cp /etc/puppet/modules/site_mcollective/site_mcollective.yaml /etc/puppet/yaml

Modify/Create /etc/puppet/hiera.yaml.
Include site_mcollective as a :hierarchy: source.

  - yaml
  :datadir: /etc/puppet/yaml
  - site_mcollective
  - common

Restart apache2.
You will need to restart the puppetmaster to read the hirea.yaml.

service apache2 restart

Modify /etc/puppet/yaml/site_mcollective.yaml.
Change the values to suit your site.

  • middleware_hosts: a list of your puppetmaster/mcollective servers.
  • activemq_password: Change it to something long and random.
  • activemq_admin_password: Change it to something long and random.
  • ssl_server_public: Change cert file name to your server name.
  • ssl_server_private: Change key file name to your server name.
  • users: a list of your users that can run mco. These users must already have accounts.
Copy the Server Certificates and Keys.

MCollective can use the puppetmaster SSL certificate authority for its SSL certificates. We will copy the files from the puppetmaster directories into the site_mcollective module so they can be distributed to the appropriate places.

puppetmaster CA Certificate.

cp /var/lib/puppet.example.com/ssl/certs/ca.pem /etc/puppet/modules/site_mcollective/files/server/certs/

puppetmaster Server Certificate and Key.
The key is only readable by root, you will need to change the group ownership to puppet so the puppetmaster process can read it.

cp /var/lib/puppet.example.com/ssl/certs/puppet.example.com.pem /etc/puppet/modules/site_mcollective/files/server/certs/
cp /var/lib/puppet.example.com/ssl/private_keys/puppet.example.com.pem /etc/puppet/modules/site_mcollective/files/server/keys/
chgrp puppet /etc/puppet/modules/site_mcollective/files/server/keys/puppet.example.com.pem 
Adding Users.

We need to generate a SSL certificate and key for each of our users. Puppet has a really simple interface to do this. I’m using the ubuntu user as an example.

Generate a Certificate and Key.

puppet cert generate ubuntu

Copy User Certificate and Key.
Again the user keys are only readable by root, you will need to change the group ownership to puppet so the puppetmaster process can read it.

cp /var/lib/puppet.example.com/ssl/certs/ubuntu.pem /etc/puppet/modules/site_mcollective/files/user/certs/
cp /var/lib/puppet.example.com/ssl/private_keys/ubuntu.pem /etc/puppet/modules/site_mcollective/files/user/keys/
chgrp puppet /etc/puppet/modules/site_mcollective/files/user/keys/ubuntu.pem
Install Middleware Server.

Now we are going to use puppet to install the middleware server, agent and mco on the puppetmaster instance.

Create/Modify your site.pp.
Assign the site_mcollective class to your puppetmaster. Instead of assigning classes to traditional node definitions, I’m using the site_role fact I created in the user-data when I built the instance. See Part 1 for details on setting up the roles with user data or simple shell script. The site_mcollective cass has 3 install_type‘s available.

agent – Default. Just the agent portion, install this on most instances.
client – The mco software and the agent. Install this the instances you want to run commands from.
middleware – The “server”. ActiveMQ, agent and the client. I’m adding site_mcollective with install_type => 'middleware' on my instances with the puppetmaster role.

case $::site_role {
    puppetmaster: {
        class { 'site_mcollective': 
            install_type => 'middleware', 
    default: {

Run puppet agent.
Now run the puppet agent on your master instance. You should see a whole bunch of actions happen. Hopefully it’s all green output.

puppet agent -t

Test MCollective.
Now login as the user you set up and run mco ping. If everything setup correctly you should see a list of instances running the agent.

ubuntu@puppet01:~$ mco ping
puppet01                                 time=61.73 ms

---- ping statistics ----
1 replies max: 61.73 min: 61.73 avg: 61.73 

Hurrah it’s working!

Install on Additional Agents.

I’m adding site_mcollective on my instances with the mrfancypantsapp role. Since install_type => 'agent' is the default, you don’t need to specify it.

case $::site_role {
    puppetmaster: {
        class { 'site_mcollective': 
            install_type => 'middleware', 
    mrfancypantsapp: {
        class { 'site_mcollective': }
    default: {

Run puppet agent.
Wait for the automatic puppet agent run or trigger it manually.

puppet agent -t

Test MCollective.
Run mco ping from your puppetmaster. You should now see additional hosts.

ubuntu@puppet01:~$ mco ping
puppet01                                 time=61.73 ms
mrfancypantsapp-prd-ec2-111-111-111-111  time=75.33 ms
mrfancypantsapp-prd-ec2-111-111-111-112  time=80.31 ms
mrfancypantsapp-stg-ec2-111-111-111-231  time=63.34 ms

---- ping statistics ----
4 replies max: 80.31 min: 61.73 avg: 70.18 

Now we can use the site_role fact to run commands on servers that are only in that role.

ubuntu@puppet01:~$ mco ping -F site_role=mrfancypantsapp
mrfancypantsapp-prd-ec2-111-111-111-111     time=75.33 ms
mrfancypantsapp-prd-ec2-111-111-111-112     time=80.31 ms
mrfancypantsapp-stg-ec2-111-111-111-231     time=63.34 ms

---- ping statistics ----
3 replies max: 80.31 min: 63.34 avg: 72.99
Install Client (mco).

I’m adding site_mcollective install_type => 'client' on my instances with the secure_jumppoint role. From these instances I can run the mco command to send commands to other servers.

case $::site_role {
    puppetmaster: {
        class { 'site_mcollective': 
            install_type => 'middleware', 
    mrfancypantsapp: {
        class { 'site_mcollective': }
    secure_jumppoint: {
        class { 'site_mcollective': 
            install_type => 'client',
    default: {

Run puppet agent.
Wait for the automatic puppet agent run or trigger it manually.

puppet agent -t

Now you can log into the secure_jumppoint instance and run mco commands.

Next Steps.

Now you have Puppet and MCollective working, “what do I do with it?” You will just have to see the next post – Part 4 for working with Puppet/MCollective, tips and some of my best practices.

Role based Puppet/MCollective in EC2 – Part 1 Roles

This is the first in a short series of posts that will take you through building simple role based infrastructure in EC2 using Puppet for configuration and MCollective for orchestration.

What do you mean “role based”?

When you are talking AWS EC2 or any other “cloud infrastructure” traditional ways of naming and organizing systems are just not flexible enough. When autoscaling, instances come and instances go. Your 4 web servers may, at any moment turn into 6 servers. When you scale back down it might not be the original 4. The EC2 assigned DNS like ec2-54-86-7-157.compute-1.amazonaws.com really don’t tell you anything about what that server does. To solve this problem, I create a couple of simple definitions (in this case Puppet “facts”) that I use to organize the systems in a human way. With tools like Puppet I can deploy my configuration and code to servers in various roles. MCollective allows me to run commands and manage the systems simultaneously by role. It doesn’t matter if it’s 4 servers or 100.

A bit of background.

Puppet and MCollective use a program called Facter to generate “facts” about a system. Facter runs on the agent systems during the Puppet agent run. These facts end up as top level variables in Puppet. For example $::hostname is facter hostname.

Assigning roles.

Now we need a way to assign roles to the instances when we boot them. To keep it simple, I’m using inline yaml in the user-data to create some basic “facts” at build time. Using yaml makes it easy just populate the user-data box in the AWS Management GUI, AWS CLI or for use in more advanced automation like CloudFormation. Here is what I include in the user-data:

  • env: Instance environment – Examples: dev, qa, stg or prd
  • role: Instance role – Examples: www, api, mongodb

{env: prd, role: mrfancypantsapp}

This simple python script reads the user-data (provided to the instance by an AWS magic url) and spits out key=value pairs for Facter so we can use these values in Puppet. I namespace these values with site=’mysite’ so they don’t stomp on values set by other data sources. Just place this script in the

/etc/facter/facts.d directory and Facter will process it on each run. I include this script in our custom AMI builds so this works at boot.

#!/usr/bin/env python
site = "site" # change this to your own site

import yaml
import subprocess

userData = subprocess.check_output('/usr/bin/ec2metadata --user-data', shell=True)
facts = yaml.load(userData)
for k in facts.iterkeys():
    print site + "_" + k + "=" + facts[k]

These facts will now be top level variables in Puppet. You can see them and other things Puppet/Facter thinks are important for identifying your system by running facter on the agent host. Using this simple yaml makes it easy to add more facts at build time.

But wait, I have a bunch of systems already running.

You can’t modify user-data on running instances, but all Facter wants is something that spits out key=value pairs. For legacy systems I use a simple shell script to echo values. Place this script in the /etc/facter/facts.d directory.

echo site_env=prd
echo site_role=mrfancypantsapp