KubeDay Australia 2024

Istio, one of the most popular service meshes, is widely used by many companies. Service meshes simplify observability, traffic management, security, and policy on Kubernetes. While they offer significant benefits, day-to-day operations like upgrades can be challenging. These upgrades require active monitoring during the process. GoTo Financial, for instance, took more than 45 days to upgrade 60+ clusters. This talk will share their journey of building an open-source, opinionated automation solution to simplify the Istio service mesh upgrade process. This solution has shortened upgrade time to 14 days, reduced active monitoring, and frees up valuable engineering resources and minimized downtime risks.

There are many use cases for scheduling in the world, and the maintainers regard extendability as crucial. Kube-scheduler has two major ways to extend it, webhook and Go SDK. However, these come with drawbacks; performance/extendability issues with webhook, and the need to rebuild and replace schedulers with Go SDK. The proposed talk seeks to introduce our new solution to these challenges - a WebAssembly based extension. We’ll begin with a story about why custom schedulers are needed, and how existing extensions work. Then, we’ll get into why architecturally WebAssembly is a great fit for plugins, while also covering its challenges, such as sandbox limitations and programming concerns. This session will be delivered by Kensei, SIG-Scheduling core maintainer who founded the wasm extension. When you leave this session, you will know first-hand how a complex WebAssembly project looks and feels, with concrete details of how WebAssembly fits into infrastructure extensibility in general.

With so many options available, choosing and architecting CloudNative Solutions is a challenging process. Historically, focus has been placed on the solutions that best address specific use-cases and how these fit into the desired overall enterprise ecosystem. Experimenting is not an uncommon practice, especially where the upstream or downstream dependencies are multi-cloud or hybrid.

So how can we efficiently and effectively design our solutions to optimize for fit, reliability, scalability and stability across our complex distributed systems?

In this session, I will cover some key design use-cases, challenges and how Observability, OpenTelemetry and AI/ML is evolving to provide streamlined insights for practitioners to architect for complexity.

In this talk, Cail will dive into an incident that occurred in Octopus Deploy's internal build platform, where thousands of ephemeral test environment pods jammed up their test cluster for seemingly no reason at all. We'll talk about incident analysis process, the specific contributing factors identified, and the ongoing challenges of the work involved in providing internal tooling.

Kubernetes is a defacto choice of resource manager for different types of requirements like running your microservices, CI/CD pipelines, Analytical applications, etc. Each of these requirements has specific challenges and as we address these challenges some new design patterns evolve which act as templates for many to use these learning and effectively run your workloads. Running AI/ML and Bigdata engines like Spark, Flink, Hive, Tez, Flume, etc on Kubernetes has presented multiple challenges with respect to handling container placement, Config management, confidential data management, enforcement of governance policies, Metadata management, Shuffle data management, Logging and monitoring, Autoscaling, Sharing data with other containers, Sharing libraries across different containers and many more. The Agenda of this talk is to discuss these challenges and how different design patterns have evolved to address them.

As organizations attempt to build internal developer platform s to standardize the way they deploy and manage cloud-native applications, they struggle with the cost and complexity of building and operating a home-grown solution. Radius is an open-source framework that enables developers and operators to build, deploy and operate on cloud native applications across public and private clouds. In this session, I'll talk about why developers and operators might want to use Radius when building cloud-native applications, go into the architecture of Radius, compare it other common tools and frameworks for managing cloud native apps, and show how you can build cloud-native applications using Radius. By the end of this session, you'll have an understanding of what Radius is and how you can use it to build and manage your cloud native apps.

Ethics in Cloud Native Development: Balancing Innovation with Responsibility Innovation is key to maintaining a competitive edge and nowhere is that truer than in cloud native development. However, as developers push the boundaries of what technology can achieve, it's crucial to pause and consider the ethical implications of these advancements. This talk will explore the intersection of ethics and cloud native development, focusing on three critical areas: data privacy, environmental impact, and social implications.

As organizations scale their cloud adoption, they continue to find bottlenecks in their ability to deploy new workloads due to the shear number of resources centralized teams must manage. Developers at these organizations must traverse through multiple workflows—stitching together outputs from multiple pipelines—to deploy their applications. These delays result in new workloads or features taking up to 6-12 months to launch into production. Attempting to remove the bottleneck of centralized teams and empower developers, technical leadership is driving self-service automation initiatives commonly referred to as platform engineering. This presentation covers why platform engineering is trending, why new tools and methodologies are required, real-world customer use-cases, and how to get started. Additionally we’ll demo how to build a modern platform using OSS CNCF tools like Argo, Crossplane, and Gatekeeper.

SpinKube [1], an open-source project, is a revolutionary platform designed to leverage the unique advantages of WebAssembly (Wasm) for Kubernetes environments. SpinKube introduces a streamlined, efficient approach to deploying serverless applications. Complementing SpinKube's advancements, the Fermyon Platform for Kubernetes [2] can run thousands of serverless functions per node. The cold start delays are less than one millisecond. This increase in efficiency translates into up to 60% compute cost savings, providing a cost-effective solution without compromising performance. SpinKube and the Fermyon Platform represent a significant step forward in cloud computing. [1] https://www.spinkube.dev [2] https://www.fermyon.com/blog/intro-to-fermyon-platform-for-kubernetes