Non-Functional Requirements: do user stories help? [PDF][Video] Rachel Davies
Cucumber-nagios [slideshare][Video] Flapjack … rethinking monitoring for the cloud [slideshare][Video] Lindsay Holmwood
Building Agile Infrastructures with Puppet [PDF][Video] Teyo Tyree
Introducing Kanban in operations [PDF][Video] Mattias Skarin
Continuous Integration, Pipelines and Deployment [slideshare][Video] Chris Read
A private openQRM Cloud use-case for a developer team … not only for sysadmins. [ODP][PPT][Video] Matthias Rechenburg

Detail

Non-Functional Requirements: do user stories really help? (Rachel Davies)

Agile is all about closing the gap between business and software development. Agile teams work closely with a Product Owner who prioritizes user stories on their business value. So how does a team make sure they don't lose sight of “non-functional requirements”? Are user stories of any use to make infrastructure more visible in the product backlog? This talk is based on coaching experiences with many agile teams. Come and hear how these teams attempted to resolve these concerns. Find out about patterns and anti-patterns that apply to non-functional requirements in an agile world.

Cucumber-nagios + Flapjack: rethinking monitoring for the cloud (Lindsay Holmwood)

Writing checks for your monitoring system is boring. You end up writing the same checks again and again, and it can be difficult to verify behavior instead of availability. Wouldn't it be useful to have a standard library of checks you could reuse across your infrastructure? Say hello to cucumber-nagios - it lets you write reusable behavioral tests in human-readable language. As cucumber-nagios output the test results in the Nagios plugin format you can run your checks from any monitoring system that understands the format, but as you start adding more machines to your monitoring system you're going to notice slowdowns and reliability problems. Enter Flapjack, a scalable and distributed monitoring system. It natively talks the Nagios plugin format, and can easily be scaled from 1 server to 1000. Flapjack aims to be simple to set up, configure, and maintain, and easily scales from a single host to multiple. Lindsay will be covering how to get up and running with both cucumber-nagios + Flapjack, writing tests for your web apps, and why it's important to test the behavior (and not just the availability) of your production web apps.

Building Agile Infrastructures with Puppet (Teyo Tyree)

The reality of cloud computing and tools like Puppet have enabled systems architects to build agile infrastructures that can nimbly react to changing business needs. We will review pragmatic approaches and guidelines to developing agile infrastructures using Puppet while exploring analogs to software development approaches in general.

Introducing Kanban in operations (Mattias Skarin)

Software development teams have had Agile going for some time. But how can operations derive from the same benefits? Or, can we do better? These were questions we faced at a game operations department one year ago. This talk is a “from the trenches” report of doing it, including how we wrestled with dismantling the “we vs. them” view of development teams and operations.

Continuous Integration, Pipelines and Deployment (Chris Read)

When Continuous Integration grows within organizations, Build Pipelines can help to manage the workflow to get software through the different checkpoints to get applications to production. This process can further evolve into Continuous Deployment. A side effect of this, is that the management of the CI infrastructure also requires an increased involvement of sysadmins and operations.

… not only for sysadmins (Matthias Rechenburg) A private openQRM Cloud use-case for a developer team

This presentation deals with how to achieve an agile, flexible and automated IT environment based on an example implementation of a private openQRM Cloud for a developer team. Implementing new features and fixing bugs are the two most important tasks for a developer team. While implementing new features always requires adaption to different operating system versions and system architectures also the second work area of fixing bugs needs lots of testing with various custom setups. Lots of time in the development process is misspend just to prepare and provide a powerful testing environment for the developer team. The performance of the test-bed is actually proportional to the quality and robustness of the developed software. The presentation focus on the different aspects of private Cloud Computing with openQRM and how system administrators and especially developers can take advantages of that.

Speed up and automate custom provisioning via a private openQRM Cloud

The openQRM Cloud provides a standardized, flexible and extensible "request system" mechanism to the development and QA teams so they can simple and fast get systems according their needs. Those systems can be fresh and clean, known-to-work testing systems but openQRM provides also the capability to easily exchange snapshots of development-, QA-, or even production-servers during runtime without affecting the origin service. This means e.g. the development and the QA team can deploy a 1-to-1 copy of the production environment as their test-bed and even exchange snapshots of specific servers for sanity checks. Server snapshots in openQRM are also useful for keeping a versioned history of specific server-images to e.g. compare application performance between different versions.

Automate Configuration Management

The Puppet integration as an additional plugin in openQRM allows further to fully automate the deployment process of the application stack on the requested Cloud Appliances. This programmable configuration layer allows system-administrators, QA and developers to "develop" their IT infra-structure via the well defined puppet-language. Via the openQRM Cloud WebService and SOAP API Aspect 1) can be fully automated and integrated with existing business processes. Furthermore openQRM is designed with a fully plug-able architecture so any third-party component can be integrated easily.
  • you do not need to be a sysadmin to integrate a third-party component
  • you do not need to be a developer to integrate a third-party component
  • ... but it is good to be a bit of both