Here you can find all videos and slides of the OSDC 2012:

Kristian Köhntopp | 8 Rollouts per Day (EN)

Booking.com lives literally agility – we rollout new code hourly every day or even more often. The primary objective of this development is the velocity- realized with a radical mode of thought and some unusual concepts.

Dr. Hendrik Schöttle | Data Protection, Software Licences and other Legal Issues in the Cloud (EN)

Cloud computing and data protection are a bad match at first sight: On the one hand distributed data storage in an unknown IT infrastructure and on the other hand the requirement of a controlled data processing in a known environment. Add web services and you have issues regarding (software) licensing as well.

The talk will give an overview of legal hurdles that are to be overcome in Germany when dealing with cloud computing.

Mark Burgess | CFEngine and why the future of configuration management is knowledge (EN)

The 21st century is increasingly about the transformation from an industrial society to an information based society. This transformation is also referred to as the "Third Wave" of western civilization.

As the demand for agility and diversity grows in the competitive commercial arena, the ability to manage variability in IT infrastructure becomes critical. This goes far beyond current ideas about the Cloud.

CFEngine 3 is a modern technology for IT infrastructure that employs a number of the techniques of knowledge management to support diversity cheaply. In this talk, Mark Burgess, the creator of CFEngine, explains the challenge of the Third Wave and what CFEngine is doing to tackle it.

Kris Buytaert | Devops and Open Source (EN)

For the past couple of years everybody from your cat to the Gartner analysts have been talking about devops, but what is this devops thing. Why does it matter .. and what does it have to do with Open Source?

We've come a long way since introducing new ideas in server automation and deployment, and also in creating a culture of collaboration between the traditional silos in organizations. But how does this impact the traditional sysadmin world? Are we all a DevOps now? Does a DevOps person = sysadmin 2.0? Will DevOps put us out of a job? I will give a brief overview of how culture, workflow, and behavior have evolved.

After evaluating the past and the present, I will talk about the future, identifying technical gaps in monitoring, packaging, and data collection and identifying emerging human, organizational evolutions.

Jochen Lillich | Operations mit Kanban (DE)

Kanban is a more and more widely used method to visualize and manage the current processes of software development and IT operations. This method can be integrated into existing processes and increases their agility. In this lecture I will explain what Kanban exactly is, the advantages of this methods and what potential impacts it could have on the culture and the processes of a team.

Thomas Gelf | Datacenter aus der Puppenkiste (DE)

The dry subject configuration management is experiencing a new revival, especially now, in times where everything is virtualized and integrated in the cloud. Not least, also thanks to Puppet.

The aim of this presentation is to introduce step by step the description language with the help of a comprehensive practical example. While the introduced scenario slowly continues to grow, this presentation will help you to better understand, whether and for what Puppet should be applied in your own data center.  Beside typical configuration tasks, you will learn more about flows and tasks from deployment to testing as well as recommendations and ideas for rollouts and rollbacks of complete server environments to round off the subject.

Andreas Schmidt | Aufbau und Einsatz von Marionette Collective (DE)

The Marionette Collective is an upgradeable server orchestration framework released under the Apache 2.0 license. It has its origins – as well as the popular configuration tool Puppet – from the DevOps forge Puppet Labs in Portland, USA. MCollective has been developed to carry out simultaneously several administration tasks in distributed environments. It contains a simple but efficient system of ruby-based agents, embedded in an asynchronous messaging infrastructure (e.g. Apache ActiveMQ or RabittMQ), providing RPC services. Advantageous skills of the messaging infrastructure as clustering and performance will thereby be inherited. The independence of a host naming convention to verify a server identity is especially in virtualized environments a novel attribute. The advantages over SSH-based approaches are the indirect addressing of the target systems through facts, the high speed thanks to the parallel execution of tasks and the ability to operate in several distributed locations such as in cloud environments.

Because of this intelligent combination of these new features you have the ability to administer a server landscape regardless of its size. Monitoring and reporting jobs, continuously incoming administration task but also the automatic deployment of entire stack applications, are practical examples of use. The application in conjunction with Puppet is a powerful tool combination for the administration of huge environments. This presentation is primarily for administrators and for people interested in DevOps. It will show the advantages, the application areas and the concrete deployment of MCollective with the help of examples.

Maciej Dobrzanski | The grass is always greener on the other side (EN)

When people are stuck on a problem with their system or application, they often fail to make a cool-headed analysis in order to identify what the problem actually is or what the next steps towards its resolution should be. Instead, they start running blind and making random judgments about some technology used in the implementation and how something else in its place would work a lot better. In many cases the behavior is fueled by insufficient skills or experience, lack of relevant data on the problem, bad advices given by other people who do not even know anything about the specifics of the project, or finally articles, discussions and blog posts on the Internet that are commonly just part of a trend, fashion or even a holy war. The session will aim to highlight this frequent problem, discuss some of the dangers of making decisions without actually understanding the problem first, and provide guidance on how to prepare for and approach problems.

Ross Lawley | Building a first application on MongoDB (EN)

MongoDB – from "humongous" – is an open source, non-relational, document-oriented database. Trading off a few traditional features of databases (notably joins and transactions) in order to achieve much better performance, MongoDB is fast, scalable, and designed for web development. The goal of the MongoDB project is to bridge the gap between key-value stores (which are fast and highly scalable) and traditional RDBMS systems (which provide rich queries and deep functionality).

This talk will introduce the features of MongoDB by walking through how one can building a simple location-based application using MongoDB. The talk will cover the basics of MongoDB's document model, query language, map-reduce framework and deployment architecture.

Alexey Kopytov | Taking hot backups with XtraBackup (EN)

Percona XtraBackup is an open source hot backup tool for MySQL. This talk will cover basic operations such as taking full and incremental backups, restoring from backups and setting up replication slaves, as well as advanced and recently added features such as streaming and compressed backups, parallel operations and partial backups.

Erkan Yanar | MySQL Cluster (DE)

With the API for Storage Engines, MySQL has different ways to backup the data.

The NDB storage engine allows above all the high availability of the data without any loss as well as a high level of parallel accesses.

Kenny Gryp | Expert Troubleshooting: Resolving MySQL Problems Quickly (EN)

Troubleshooting MySQL doesn't need to involve trial-and-error, and it doesn't need to take a long time. You need only two things: a good process and good tools. This session will show you an approach that the speaker has used to solve many frustrating problems quickly, and opensource tools that assist this process. The tools especially include key tools from Percona Toolkit, designed to help diagnose problems that can't be observed directly.

Ross Lawley | Scaling with MongoDB (EN)

MongoDB's architecture features built-in support for horizontal scalability, and high availability through replica sets. Auto-sharding allows users to easily distribute data across many nodes. Replica sets enable automatic failover and recovery of database nodes within or across data centers. This session will provide an introduction to scaling with MongoDB by one of the developers working on the project. 

Erkan Yanar | Galera: Multi-Master Replikation für MySQL (DE)

Galera is an add-on for MySQL that allows synchronous multi-master replications (of InnoDB tables). This speech will present the setting of Galera and how often used fragile MySQL replication architectures can be replaced stablely by Galera.

Olivier Renault | Introduction to Eucalyptus (EN)

Eucalyptus is an open source solution enabling you to create on-premise private clouds, with no requirements for retooling the organization's existing IT infrastructure or need to introduce specialized hardware. Eucalyptus implements an Infrastructure as a Service (IaaS) private cloud that is accessible via an API compatible with Amazon EC2 and Amazon S3. This compatibility allows any Eucalyptus cloud to be turned into a hybrid cloud, capable of drawing compute resources from public clouds. Eucalyptus is compatible with a wealth of tools and applications that also adhere to the de facto EC2 and S3 standards.

This talk will introduce the architecture of Eucalyptus as well as describing the best practices to implement a Eucalyptus Cloud. We will finish by a quick demonstration on how to scale / burst to a public cloud.

Constantino Vázquez Blanco | OpenNebula Open Source Toolkit for DataCenter Virtualization (EN)

OpenNebula is the open-source industry standard for data center virtualization. The presentation will describe its innovative features for the comprehensive, complete management of virtualized data centers to enable on-premise IaaS clouds in existing infrastructure. OpenNebula interoperability makes cloud an evolution by leveraging existing IT assets, protecting your investments, and avoiding vendor lock-in.

Dr. Christopher Kunz | CA failures and the future of Web authentication (EN)

In 2011, a number of Certification Authorities suffered catastrophic failures which showed that the SSL CA system, a cornerstone of the secure Web, has been undermined by attackers and corporate greed. These failures and malpractices may well lead to the eventual downfall of SSL certificates as we know them. 

This talk will summarize the events which transpired last year (and continue to pop up in 2012) and show which alternatives are currently in the making. It will introduce concepts like DANE, Convergence, Sovereign Keys and show some interesting info about SSL certificates "in the wild".

Dr. Christian Winkler | Ultra-performant dynamic websites with Varnish (EN)

We describe how the use of the Varnish web cache can lead to tremendous performance gains for dynamic websites like online shops or portals. Peak request rates of several thousand pages per seconds can easily be achieved without putting too much stress on the (commodity) hardware.

Martin Gerhard Loschwitz | Die eigene Cloud mit OpenStack Essex (DE)

OpenStack is a budding solution for enterprises that want to build their own cloud. The environment provides all the features to meet the essential requirements of a digital cloud: Nova handles the administration of resources and is responsible for starting up new virtual instances, Glance manages existing operating system images and Swift provides memory storage inside the cloud. The OpenStack-Dashboard called Horizon, is the control center for administrators and furthermore allows the acces to the OpenStack user management (“keystone”).

Dr. Udo Seidel | Extremes Wolken Dateisystem!? (DE)

So-Called shared file systems have been known in the Unix / Linux-environment for a long time. The network-based and cluster data systems are thereby almost an old hat. The project XtreemFS– a part of the grid operating system XtreemOS – is a quite new but budded alternative in this field. The current version finally has some important features and is worth taking a closer look.

This presentation gives an insight into the data systems approaches, explains the architecture and describes the first steps of a XtreemFS-Cluster set-up.

Benedikt Stockebrand | Like rats on a sinking ship (EN)

By now most of the IT industry has finally realized that IPv6 is soon becoming inevitable.  But many people don't realize what will happen with IPv4 some time throughout this year: As soon as ISPs introduce Carrier Grade NAT (CGN), the resulting functional limitations won't just cause some extra work in the IT departments but will render popular business models infeasible.

The talk gives an overview what to expect during the upcoming final stage of IPv4 and the introduction of IPv6, how to handle it and why open source is so crucial during this period.

Berthold Gunreben | Implementing Converged Networks in a Virtualized Environment (EN)

Getting more speed, lower latency, lower costs, and keep redundancy for block and network devices. This is what converged networks promise.

Looking at the many advantages, it is somewhat surprising that implementations still lack behind. The services team at SUSE Linux picked up the task, and implemented a virtualization environment based on KVM that runs critical internal services. We want to shed some light on the expectations and pitfalls we encountered during implementation.