Über uns
Karlsruhe community for all things Kubernetes and CNCF-related tech.
People of all experience levels welcome, from beginner to expert. Whether you're dev, sec, and/or ops.
We will regularly host talks, discussions and networking events.
Bevorstehende Events (1)
Alles ansehen- 🐑🐑 SCALING CONTROLLERS 🔬 AUTO INSTRUMENTATION - Cloud Native Karlsruhe AprilCodesphere GmbH, Karlsruhe
We're hopping out of the easter break by bringin home a slice of KubeCon/Cloud Native Con London 🎡 (Replacing the terrible conference food with many good slices of pizza 🍕)
Talk #1: 🐑🐑 Beyond the Limits: Scaling Kubernetes Controllers Horizontally
Do your Kubernetes controllers struggle to keep up with the demands of your growing infrastructure? As your clusters scale, traditional controller setups face increasing challenges, leading to slow reconciliation times, impacting application performance and overall cluster stability.
This session introduces sharding for Kubernetes controllers as a groundbreaking solution. By horizontally scaling controller workloads across multiple instances, it significantly improves scalability and addresses the inherent limitations of traditional leader election mechanisms.
In this session, we'll dive deep into the technical details of applying proven sharding mechanism from distributed databases to effectively partition controller workloads. We'll explore the underlying concepts and how to implement sharding in your own Kubernetes controllers.
Join us to learn how to overcome the scalability challenges of your Kubernetes controllers and unlock the full potential of your infrastructure.
Tim Ebert loves designing, developing, and operating cloud native systems at STACKIT. He is knee-deep in managing infrastructure and Kubernetes clusters themselves using Kubernetes operators. Tim is a core developer of Gardener, an open source project for managing Kubernetes clusters at scale. Before joining the STACKIT Kubernetes Engine team, he was part of the Gardener team at SAP.
Talk #2: 🔬 The Art and Craft of No-touch Instrumentation
OpenTelemetry automatic instrumentations greatly help with achieving a high baseline of observability for applications. But applying them in containerised environments still requires manual, error-prone intervention by users. LD_PRELOAD-based injection is an advanced technique that has been used by commercial vendors for a decade or more, is by now very well understood, and does not require access to the Kernel like eBPF does (which is a huge challenge in managed compute environments), or any increase in permissions other than what the host runtime has.
In this talk, we go in depth into how this magic works across various runtimes, and how it could benefit the OpenTelemetry community at large.
Michele is the Founding Engineer and Head of Product at dash0, an Open Telemetry client. He has been living breathing observability for the past decade. It all started when his product owner asked "Why does this user claim to be logged into someone else's account?"
AGENDA
18:30: Arriving and networking
18:45: Talk #1
19:15: Break, discussion, and pizza
19:45: Talk #2
20:45: Time for discussion and networkingLooking forward to seeing you 💙
