Recently I had the pleasure of attending the first annual PuppetConf, put on by our local friends at PuppetLabs here in Portland. The speakers were top notch as expected, representing companies such as Google, Facebook, VMWare, Amazon, etc… and of course Janrain. Most of the talks were presented as “here is how we use Puppet at our company”, which was great and actually surprising to hear the different ways people use Puppet to make their lives easier and more efficient.
One of my favorite talks was by Digant Kasundra from Stanford University. He went over the “Puppet is great and here is why”, like most people, but then went on to include pitfalls and how they learned what not to do with Puppet and why. He also explained Classes vs. Defined Types, along with why they seem interchangeable but should be used for very different things. You can find a bit of this and other knowledge from him in one of the Puppet Best Practices documents.
During the course of the two days I had the pleasure to meet a few of the actual PuppetLabs engineers who were very friendly and helpful with questions about their software. One in particular was around all the changes coming in 2.8 and how we at Janrain can better prepare for the coming upgrade was very insightful. One of the many changes coming in 2.8 means that dynamic variable look-ups are being deprecated, and we’re now to use fully-qualified variable names. One of the PuppetLabs guys took the extra step to point me in the right direction, and now our Puppet logs are in great shape.
Over all it was a great experience to rub elbows with other people in the Puppet community. I love my job here at Janrain, and this conference made me love Puppet too. When I started at Janrain, I had very little exposure to Puppet and it’s Ruby-ish looking code. Five months later with much help from Janrain Operations “Puppet guru” James Loope, and two days at PuppetConf, I get it.
Puppet allows us the peace of mind to keep things automated and consistent across our whole infrastructure. It’s also nice to have the ability to provision new instances in minutes and scale if needed on a moments notice.
Thanks to all the people that made this conference happen and run so smoothly – hope to see you all next year.