Custom External Nodes

As a starter example, last year we announced Puppet Dashboard, a GUI that allows you to gain an easier view and access into your Puppet infrastructure.

The “node classifier” in Puppet Dashboard is built on top of a powerful system Puppet has had for a long time, called External Nodes, which tells a given host what Puppet classes and variables should be assigned to it, based on the return calls of a command execution where the hostname of the system is an argument to the command. The command just returns a simple set of data in YAML format on standard out. This is totally dynamic and perhaps one of the easiest puppet extensions to write.

While also enabling Dashboard, External Nodes gives you the choice to also use your own in-house tool to provide this information, for instance, that might integrate with your existing LDAP infrastructure or another database. It’s also how the Puppet Integration in apps like Cobbler work.

Custom Types And Providers

Another powerful feature is the ability to write custom types. While Puppet ships with a very large library of standard resource types, it’s quite easy to write your own to manage a resource stock Puppet does not manage out of the box. For example, recently we’ve developed a type for Managing LVM volumes. If you need assistance writing types, it’s suggested that you read the source code for some of the existing types. The LVM one is a great example. If you are still stuck, we can help.

Custom Facts

Similarly, if you need to evaluate conditional expressions or source template variables that Puppet doesn’t provide out of the box using Facter, Facter can be trivially extended by writing Custom Facts. For instance, you may wish to write a custom fact that provides some added information about hardware inventory, or perhaps information about specific in-house software.

Custom Functions

For even more power, it’s possible to write custom functions. While Facts are evaluated on the managed-nodes, functions are evaluated server side. In this way, it is possible to write a function that looks up data from a file. One such example is extlookup. Even if you wanted to write most of your data in Puppet language, including some bits from external sources is fairly simple.

Language Options

It is true that custom facts and functions currently require writing Ruby code, but External Nodes can be written in any language that can produce output in simple YAML. We’re also considering a way to allow custom facts to be written as simple shell scripts dropped into a conf.d type directory. While traditionally manifests have been required to be written in Puppet Language, work on our Ruby DSL will allow manifests to be described in native Ruby. Note that these manifests are intended to be run on the central server, not on the clients, and the result of running them is the creation of a Puppet catalog. If you’re doing things on the clients and wish to use Ruby, writing a type and provider is the way to go and that’s been doable for some time.

A Pluggable, Open System

The fundamental takeaway is this — Puppet is an flexible open system, and your data and needs should decide how to drive your infrastructure. Puppet provides many ways to make that possible, and in the future, you’ll see this continue to grow and evolve. Rather than having an giant enterprise tool that provides a set work flow and locks in your data, we want Puppet to be a system that can easily adapt to your particular needs.

–Michael DeHaan

Our Founder and CEO Luke Kanies will be speaking remotely as well, so if you are in India this weekend you should definitely plan on attending.

The day’s goals are:

• to give newcomers the opportunity to learn Puppet in a collaborative environment (with a two-hour guided exercise on EC2 machines that we will provide)
• to apply that knowledge in a one-day competition immediately following the exercise, on the same machines (or on other machines if you like), or apply the principles in the systems automation tool of your choice, i.e. Puppet, cfengine, chef, lcfg, bcfg2
• develop the community of like-minded senior operations professionals focused on furthering the promise of systems automation – and the profession

We hope you can check it out and let us know how it went!

Things to include in your proposal:

• Presentation title: a punch line describing your talk
• Abstract: a summary of your talk (around 70 words)
• Keywords: a relevant set of keywords
• Benefit for the audience: what do you want the audience to take away from the session
• Targeted audience: technical, manager, business, …
• Targeted experience level: beginner, intermediate, expert
• Your contact details: name, email address, company name
• One line Bio: explaining your background, experience level

Looking forward to hearing your great stories! Send your proposal to submissions@puppetcamp.org

Want to register for Puppetcamp Europe? Submit registration here:

Register for Puppetcamp Europe

More details will be coming soon regarding speakers and agenda. The conference will follow Open Space principals and will be very interactive for all attendees. We look forward to seeing you there.

Thanks to our sponsor Optiver for helping make this happen.

The article gives some great insights into the challenges of running an online game and how they interact with Facebook. At the end of the post Luke wrote about how he manages their web farm:

To help manage and monitor FarmVille’s web farm, we utilize a number of open source monitoring and management tools. We use nagios for alerting, munin for monitoring, and puppet for configuration. We heavily utilize internal stats systems to track performance of the services the application uses, such as Facebook, DB, and Memcache. Additionally, when we see performance degradation, we profile a request’s IO events on a sampled basis.

It’s good to see Puppet being used in some of the world’s most demanding environment where efficiency, reliability and predictability are essential.

Location & Dates

• London, UK – March 29-April 2
• New York, NY – April 12-16
• Nuremberg, Germany (hosted by Netways) – April 20-22

Becoming a Puppet Master – 3 Days

Puppet Training consists of 3 days of hands-on training performed by a Reductive Labs Puppet professional. Attendees will be taught the principles and best practices of Puppet in a series of lectures and labs.This training is ideal for those who want a Puppet jumpstart. Newer members at an organization already using Puppet, or experienced sysadmins wanting to bring Puppet into their team will get everything they need to deploy solutions.

Topics covered include:

• Configuring Puppet and Puppetmaster
• Resource Types and the Resource Abstraction Layer
• Virtual Resources, Exported Resources and Stored Configs
• Meta-parameters, Dependencies and Events
• Classes, Modules and Definitions
• Tags and Environments
• Puppet Language Patterns and Best Practices

Puppet Developer Curriculum – 2 Days (NY & London Only)

This is an advanced course for those Puppet users who are interested in developing skills and learning best practices for creating their own custom Resource Types and Modules.

• Introduction to Ruby for Puppet
• Advanced Function and Fact development
• Resource Type and Provider development
• Testing practices and RSpec for Puppet

Looking forward to seeing you there!

The person who submitted the winning entry will receive a BugBundle from Buglabs. It’s in your hands to determine who wins.

Michael brings a strong background in open source software and community development to Reductive Labs. As the creator and community lead for Cobbler at Red Hat, Michael oversaw the growth of Cobbler, which is now used in thousands of datacenters across the world — including in the financial industry, hosting companies, render farms, grids, and universities. Cobbler is well known in the Enterprise Linux and Fedora space as the OS provisioning tool of choice for rapid deployment in medium to large-scale environments. It is very frequently used in conjunction with Puppet to maximize flexibility and efficiency in rollouts of new machines, whether physical or virtual.

In addition to the growth of the use of Cobbler, Michael guided tremendous growth of the contributing community. The Cobbler project has had over 80 code contributors and many more community members that assist with testing, advocacy, ideas, and support.

Michael is a published contributor to Red Hat Magazine and has presented at such events as Red Hat Summit, Red Hat Cloud Forum, the Fedora Users and Developers Conference, HP Tech Forum, and local software events. Michael is also a contributor to over 50 US patent applications in the area of configuration management and datacenter automation.

“Puppet has accomplished something few open source projects achieve — not only has Reductive Labs built a best of breed configuration management platform, it also has created a large and vibrant community of users and contributors that help guide its development,” said Michael about why he was joining Reductive Labs. “The future for Puppet’s ecosystem is extremely promising, and I look forward to helping it evolve and grow further in the years to come. Whether we are talking about cloud architectures, virtualization, grid, or classical server rollouts — as datacenter application deployment gets more complex, Puppet is around to help make the complicated simple and the impossible possible. For me, this is really one of the most exciting spaces in technology to be in, because not only can you help users all over the world solve their management challenges, but you also get to drive the forefront of computing.”

“We couldn’t be happier to add someone with the open source development and community credentials of Michael to our team,” said Luke Kanies, Founder and creator of Reductive Labs and Puppet. “Michael will help us identify opportunities to enhance the value of Puppet and engage further with our already strong and passionate community. We look forward to Michael helping guide the future of Puppet.”

Martin went on to say:

“With Puppet I don’t have to worry anymore. Once I have written and deployed the profiles I can count on Puppet ensuring timely updates and consistent configurations across all my systems. More than anything Puppet saves me time that I simply can’t afford to lose in supporting my data centers.”

Among the key benefits that Sun Microsystems has gained from using Puppet are:

• System consistency
• Improved effeciency
• Met compliance standards
• Increased visibility

Get the full details and download the case study by clicking below

Sun Microsystems Case Study

In his words, the reason he ultimately decided on Puppet hinged on its ability to handle more complex tasks:

Puppet seems to handle more complex tasks, at the possible expense of more complex configuration, at least initially, but it can then handle “higher-level” funtions, like package management. This will be important to us if we are doing cross-platform (Linux as well as FreeBSD) management.

Many of the people we speak with are comparing Puppet with CFEngine and ultimately looking at the advantages they can gain from a more powerful tool. In part, Puppet was developed by Luke Kanies to address many of the shortcomings of CFEngine and to provide better tools to system administrators.