Here you can find all videos and slides of the Nuremberg Camp 2012:
James Turnbull | The current state of Puppet
James Fryman | Refactoring Puppet: Leveling up to Data Driven Infrastructure
As your Puppet Infrastructure grows, so does the complexity of the Puppet codebase. Resource relationships become very difficult to trace and follow, catalog compilation times grow, spec testing (if you're doing that) takes a long time, and any change to the Puppet code could introduce any number of regressions to the Infrastructure.
Fortunately, the software industry has been working on this problem with software development for many years. Now, it is time to start applying these practices to our own Puppet code and start This talk will discuss how to refactor Puppet code in a controlled and organized method. The goals of such refactoring will allow ease of maintenance, speedups in spec testing, and the ability to use data-driven infrastructure. This talk will discuss tearing into codebases, looking for code smells, and how to fix them. In addition, this talk will focus on how to organize modules to take advantage of Data Drive Infrastructure, allowing fully dynamic infrastructures.
Pascal Hahn | Managing Puppet with Jenkins and GIT
Nokia's Location and Commerce division is running a large set of diverse applications in their datacenters worldwide. In order to enable engineers to deploy their applications themselves and not have to worry about how the puppet-setup works, we built a multi-tenant environment that enables rapid deployment on a large scale while requiring a minimal amount of knowledge about the environments that code is being deployed in.
One of the core pieces of the platform is a fully-automated build-pipeline that takes in puppet-code, verifies it in several ways, packages it up and distributes it to puppetmasters worldwide with no manual steps involved.
This talk will sched some light on challenges we faced along the way and how they were overcome.
Olivier Renault | Using puppet to build and manage smart private and hybrid clouds
Olivier Renault will explain the Eucalyptus module for Puppet and demonstrate how this module can be used to automate the deployment of a consistent, global private cloud across more than 15 datacenters on 3 continents. The second part of the Eucalyptus workshop is an overview on how to use Puppet to manage an application deployed in an hybrid cloud, in between a Eucalyptus private cloud and Amazon Web Services. Of course, we will leave ample room for Q & A
Peter Simon and Uwe Stuehler | Agile Operations with Puppet - How to ensure feedback loops at every level
Despite the benefits of working with Puppet, it can still feel like a drag on productivity initially, especially for small changes. We used to execute our decisions immediately and now have a relatively time-consuming process: waiting for Puppet runs during development, waiting for automated tests, peer review, staging environments, and controlled roll-out.
At mobile.de and ebay-kleinanzeigen.de we want to work fast and reliable. So in this talk we will explain our current workflow with a focus on how we increased throughput with fast feedback cycles on several levels that also help us build up shared knowledge within a bigger operations team.
Patrick Otto | Learning through mistakes and pain - One year with Puppet
Puppet is known to provide out of context a multitude of machines in the cloud or in huge environments.
But what if you want to model the entire company´s infrastructure? Begun by interne systems like DHCP and mail servers up to the virtual machines of each developer in each project….
This talk will show the issues, the mistakes, the starry-eyed assumptions and why it is fun after all.
Ken Barber | Modern Module Development and Usage
This presentation focuses on some of the more modern development techniques used for developing and managing your modules.
We'll delve into some good development and module testing practices I follow myself, and talk about the tools used for managing your modules include the newer Puppet module face.
Thomas Gelf | Talking with Puppet
Your puppet is up and running, reporting is great - however it still feels like a new alien in your grown set of management tools. It has it's own frontend, collects and stores data and doesn't show any intent to automagically join your toolchain.
Of course, the source is free and open. But as long as Ruby is not among your preferred scripting languages, this doesn't help much. You stumbled over more then one database and text files using a markup language with the worst specification you've ever seen.
This talk will try to shine some light on this parts of puppet. Even if initially it doesn't look so: it's pretty easy and makes fun. And many OSS projects have already successfully done so.