Pulling the Strings

Derek Austin is a technical project manager at Puppet Labs. In a prior life, Derek was involved with a big change initiative to bring about more continuous delivery practices, as well as a DevOps culture, to his organization. The company needed a more rapid and repeatable deployment process, and needed to reduce manual errors and friction between development and operations teams. Derek talks how migrating first to Open Source and then to Puppet Enterprise helped bring about these changes.

Derek provides a lot of insight into improving tooling, process and culture:

• How embedding QA resources with development teams can help achieve truly production-ready code at the end of a development sprint cycle.

Derek makes the point that change isn't always easy, and that sometimes a change agent has to have thick skin to be able to deal with the emotional blowback. He gives cautions about providing specific reassurances, because change is unpredictable, and you're likely to make changes repeatedly. Instead, Derek suggests providing regular feedback through retrospectives at the end of each sprint cycle, and to avoid taking a too-opinionated stance so you don't provoke conflict.

• Read Derek's blog post about migrating to Puppet Enterprise to enable continuous deployment.

In the latest edition of the Puppet Podcast, you’ll find out how to start and grow a Puppet User Group (PUG) in your city. Nick Huanca, Denver PUG organizer, Byron Miller, Austin PUG organizer and Kara Sowles, Portland PUG organizer, all sit in with me to share the tips and tricks they’ve discovered to keep a PUG lively and meeting regularly. If you’re interested in starting a PUG in your city or want to know more about our PUG program, then this is the podcast for you.

You can also check out our many other recent podcasts by visiting our podcast page or subscribing in your favorite podcast tool.

• Find a PUG near you.

Adrian Cockcroft was the cloud architect at Netflix from 2010 to 2013, and took part in the initial process of evaluating, validating and migrating Netflix services to the public cloud — an initiative that started in 2008.

Adrian has keen insights into the work culture and processes that make Netflix such a remarkably innovative organization. In this podcast, we talk about what it means to be on the bleeding edge of technology, and how to push disruptive new projects forward. You'll also hear:

• Why Netflix decided to move the company's video streaming services to a public cloud infrastructure

Adrian says to watch out for centralized control, as it can stifle innovation. Instead, he recommends you determine overall business goals from the top, and give teams the ability to take action on them by trying out new technologies, implementing prototypes, and releasing open source versions to the public for feedback. A workflow that allows testing and refinement enables everyone to become a pathfinder in their own research lab.

Enjoy the podcast!

• Hear Karl Matthias, site reliability engineer at New Relic, talk about how shipping multiple times per day changes your infrastructure management.

Theme music: "Smirking" by BenSoundBeats

We recently talked with Karl Matthias, site reliability engineer at New Relic. Karl works on the company's huge-scale database infrastructure, where over 300 billion metrics are written to disk every day.

In this podcast, Karl talks about the role of automation at New Relic, and how automation and configuration management enable consistency and reliability in the company's infrastructure, letting operations staff focus on higher-level problems, along with development engineers, and on managing capacity.

We also discuss:

• How the demand to ship multiple times a day has changed the way infrastructure is managed

One of the hardest parts of bringing about change? Karl says it's convincing your organization that there's a problem. He suggests taking time to engage your team in the conversation before jumping to the solutions that you've brainstormed. Solicit a lot of feedback and hone your argument before approaching management, so that you've built support and have identified concrete, shared solutions.

Listen in, and tell us what you think. If you have a story about making positive changes on your ops team, email [email protected].

• Hear additional conversations with Gene Kim, Kelsey Hightower, Gareth Rushgrove and others.

Theme music: "Smirking" by BenSoundBeats

In this edition of the Puppet podcast, Kara and Kent sat down with Felix Frank, Puppet power user and systems engineer, and Josh Partlow, Puppet Labs software engineer, to chat about contributing to the Puppet core projects — primarily Puppet, Facter and Hiera. Felix is one of the top contributors of all time to the Puppet mailing lists and a Featured Community Member.

Felix got started contributing to the Puppet Core projects by writing the first couple of Puppet patches for his employer. For folks new to contributing to Puppet core projects, Felix suggests, “For me, the best thing to start with was Puppet types and providers. It’s work that is quite obvious and palpable and straight-forward. You can get results fairly quickly.”

Josh agrees that the Puppet types and providers extension is the best way to get started and what he sees mostly from the community. They go on to outline the entire process of contributing to the Puppet Core, including the most important and difficult pieces.

Keep listening to hear Felix and Josh talk about how to write tests for your patches. Josh says,”Don’t be afraid to ask for help.” He goes on to give advice on the best ways to get that help.

Felix talks about his new book, Puppet Essentials which he created for folks who are used to writing software or operating servers and want to get started using Puppet quickly. He goes on to talk about how it can help you ramp up on Puppet, how his book came to fruition and his stint in blogging.

Find Felix on GitHub and check out his blog. Follow him on Twitter @felis_rex.

If you enjoyed this podcast, you should check out our podcast page for more great discussions, or subscribe in your favorite podcast tool.

Don’t forget to email us at [email protected] — we’d love to hear about future podcasts ideas or questions you have.

• Read about how to contribute to Puppet Core projects.

In this podcast installment, Rob Reynolds, software engineer at Puppet Labs, talks about some of the tools that make it easier to automate Windows environments, including Powershell, Puppet, and the package manager he wrote called Chocolatey.

This discussion is especially timely, given the release of the new Windows module pack, which includes the following Puppet Supported and Approved modules:

• puppetlabs-acl for managing Access Control Lists

Just getting started automating Windows systems? Rob recommends starting small, and adding more over time. Windows Feature is a great place to start; then, begin auditing and maintaining permissions, which will help your automation efforts in the long run.

For more information, check out the Puppet Labs documentation on Puppet on Windows as well as the Puppet Supported and Puppet Approved modules for Windows.

• Listen to more conversations with Puppet experts, customers and industry leaders.

In this edition of the Puppet Podcast, hosts Kara and Kent sat down with Gareth Rushgrove, Puppet Labs software engineer, to talk about Test Driven Development (TDD). This podcast is especially helpful to those who are new to TDD: the process of writing tests first, then writing enough code to pass the test and finally refactoring the code to make it better.

Gareth shares how he uses Test Driven Development and why he’s found it a compelling way to write code. He says, “Tests help you move quickly and not break functionality.” He encourages those new to TDD to jump in and write their first test. He explains,“Writing the first test is the hardest, but once you’ve gotten one test in, you can replicate that to different modules very easily.”

Gareth recommends reading the code of modules that have test suites, such as garethr-docker and other Puppet Approved modules and to join the puppet-users mailing list if you’re just starting out writing tests. Be sure to check out the Learn More section below for even more resources.

Want to know which tests to write, which tools and frameworks he recommends for different testing situations, how to get started with TDD, and more? Listen in!

If you enjoyed this podcast, you should check out our podcast page for more great discussions, or subscribe in your favorite podcast tool.

Don’t forget to email us at [email protected] - we’d love to hear about future podcasts ideas or questions you have.

• Gareth’s PuppetConf 2014 talk on TDD

In this edition of the Puppet podcast, hosts, Kara Sowles and Kent Bye, interview Chris Price, one of the primary engineers working on Puppet Server here at Puppet Labs. We had a compelling conversation discussing the many aspects of Puppet Server that came out in the 3.7 Puppet Enterprise release, including how it's built on Trapperkeeper and why they chose to write it in Clojure.

Puppet Server started out as weekend coding stints for Chris, and grew into a next generation replacement for the Puppet Master. Chris had the opportunity to demo Puppet Server during Puppet Labs CEO and Founder, Luke Kanies’ keynote at PuppetConf this September. He said, “It was one of the more interesting experiences that I’ve had in a long time.” Listen in to find out why!

Chris tells us about Trapperkeeper (hint: It’s not a binder from the ‘80’s) and how they decided to build it: “We spent a lot of time looking at things that exist already and found our needs were much simpler and felt we could write something in Clojure that would provide a better end user experience.” Get the goods on what Trapperkeeper makes available for Puppet Server, PuppetDB and other PE features like RBAC and node classification.

Chris gave us so much great information. You’ll need to listen to find out the rest, including answers to the following:

• How do you see the new ecosystem playing out?

Are you intrigued? Chris welcomes suggestions and bug reports from the community. Also, feel free to find him on FreeNode under cprice404.

If you enjoyed this podcast, you should check out our podcast page for more great discussions, or subscribe in your favorite podcast tool.

Don’t forget to email us at [email protected] - we’d love to hear about future podcasts that you want to hear and any questions or concerns you have.

• Check out Puppet Server on GitHub

In today's installment of Conversations with Puppet Labs, we talk with Terri Haber, a professional services engineer here at Puppet. Terri works closely with our customers, and helps set them up for success with Puppet Enterprise.

Listen in as Terri discusses:

• Key cultural characteristics, like collaboration and stakeholder buy-in, that help an organization adopt Puppet successfully

As Terri sees it, a tool alone won't cause positive changes in an organization; it takes people, and it takes cultural readiness. 

Tell us what you think of the podcast! And if you have a story about making positive changes on your ops team, please send it our way by emailing [email protected].

Theme music: "Smirking" by BenSoundBeats