Contributing Value: why SMEs should be Involved in their Software Supply Chain

The majority of CNCF contributions come from large international corporations or vendors.

At my workplace, we are a CNCF End-User, and we have a team of ~10 developers. However, we find great value in allocating time and company resources to participating in open-source communities that form part of our supply chain.

Contributing to open-source is not an act of charity; the business sees greater value in return than the raw staff cost.

This talk outlines our technical and commercial rationale for involvement, despite our modest engineering capacity. Benefits include security vulnerabilities, advocating project advancements, and using it as a learning and coaching experience for developers of all levels. However, there are also some challenges.

We have become recurring contributors to several large projects over the years and now even hold maintainer positions in a couple of them.

OKD: an Opinionated Kubernetes Distribution

The CNCF Cloud landscape is vast and interoperable. Clusters come in all shapes, sizes and configurations. Linux Distributions provide "batteries included" installations with key components chosen for you - how well does this concept apply to Kubernetes Distributions?

OKD is the upstream community of Red Hat's OpenShift Kubernetes offering. A la Fedora/CentOS vs RHEL.

OKD makes numerous decisions for you out of the box, including your Node Operating System, security constraints within Kubernetes, built-in image registries, monitoring, observability, and OLM.

Beyond the base cluster is an ecosystem of operators (OKDerators) with integrated versions of Rook, KubeVirt, ArgoCD, Istio + more. They are augmented with custom console plugins, CLI tooling and documentation to provide better dev+admin experience.

This talk will explore some of the key opinions and present live demos of both the base cluster and select OKDerators deployed in a test environment.

Help! I started a Minecraft server and now I maintain a Kubernetes distribution

This is a lightning session that explains why and how I became involved in the OKD (OpenShift Upstream) working group and was eventually appointed as co-chair, as well as the advantages that came from contributing and participating in open-source communities.

Behind the Block: Minecraft Supersized (retired)

One of the first talks I toured. How Minecraft servers "professionalised" and the tech (and logistical) challenges we faced trying to maintain extensive infrastructure (1000s of application instances, gigabits of traffic, 6TB+ of RAM) whilst still in school!

This talk is no longer available and has been superseded

Crash Course in Race Management

Talk related to race management within sailing and the key roles within committee vessels.

This is not a tech talk!

Crafting a Career in Tech - Unconventional Paths & Unexpected Sidequests

Targeted towards early careers.

Don't Dig Down: Hard Lessons Scaling Multiplayer Games

How do you take a hobbyist Minecraft server and productionise it for global scale?

This is the engineering story of CubeCraft Games, founded in 2012, now an official Minecraft partner that has welcomed over 100 million players and serves tens of thousands of concurrent players.

Unlike other UGC platforms like Roblox or Fortnite, we run all the tech and infrastructure ourselves. That means global server fleets, hundreds of software components, our own network, database clusters, orchestration systems - and yes, now even a GraphQL API gateway running on Kubernetes with microservices (buzzword bingo included).

In this talk, you’ll join us as a small group of high school students just trying to keep the server up, and follow how our approach to tech and product evolved to meet player demand. Along the way, we’ll dig into the home-grown solutions that supported that growth - the good, the bad, and the downright ugly (who needs message brokers when you can push gigabits of traffic over a private IRC network?).

In part 2 of the talk, we’re transforming into a more standardised engineering stack, replacing our “esoteric” tech stack with more standard tools like Kubernetes, Quarkus, and GraphQL. We'll look at our approach to choosing technology and then how we engage and contribute to the upstream open-source communities that we rely on.

Whether you’re into distributed systems, war stories from technical mishaps, or just curious how a small team survived the industry at scale, this is a journey from flying by the seat of our pants to building a sustainable engineering organisation.

Usually delivered in two halves