Participation in KubeCon and CloudNativeCon Europe 2024 in Paris

March 2024, Paris, France - KubeCon and CloudNativeCon Europe 2024, held in Paris, was a key event for professionals from around the world, attracting a record 12,000 attendees.

This year's conference emphasized the integration of Kubernetes with AI and cloud-native technologies, while focusing on the pressing requirements of sustainability, observability and the strategic use of edge computing.

Kubernetes on Edge Day highlighted the transformative role of Kubernetes in edge computing environments, demonstrating its scalability and flexibility for deploying applications close to data sources. Key sessions focused on integrating IoT devices with Kubernetes clusters and improving real-time data processing capabilities at the edge.

Erik Nordmark from ZEDEDA and other speakers emphasized the importance of secure, scalable Kubernetes solutions for edge deployments. Discussions included how Kubernetes facilitates the management of edge devices through optimized software deployment and robust security protocols, which is critical to the distributed nature of the Emulate project.

SNCF discussed the transformative power of edge computing by integrating cloud-native technologies such as Kubernetes, containers and microservices directly into their rail operations. This strategic deployment enables real-time data processing directly at the source - where the trains run and the data is generated. By using these technologies, SNCF is improving route management, predictive maintenance and emergency response in real time, significantly increasing operational efficiency and safety.

Given the rapid development of AI applications, an important topic at the conference was sustainability in cloud-native ecosystems. Discussions focused on optimizing AI workloads on Kubernetes to improve total cost of ownership, performance and, most importantly, energy efficiency. Techniques such as using specific processors for AI inferencing and optimizing GPU usage were highlighted to meet these requirements.

In addition, Gaulter Barbas Baptista from Deutsche Bahn presented methods that enable developers to make decisions that lead to greener IT practices.

By leveraging the scalable nature of Kubernetes, strategies such as vertical pod autoscalers and tools such as KubeCost were explored to optimize resource allocation and minimize energy consumption.

Observability was a hot topic, with numerous vendors presenting solutions for managing cloud-native workloads. The introduction of OpenTelemetry has improved the standardization of data creation and collection across systems, which is essential for implementing distributed tracing and managing edge applications.

An insightful session from Apple titled Intelligent Observability laid out the future of observability in AI and discussed how observability needs to be integrated into AI model training and inference pipelines to optimize resource utilization and performance.

The development of observability techniques, particularly through Extended Berkeley Packet Filtering (eBPF), received a lot of attention. Groundcover and Cilium presented their budget-friendly eBPF-based observability solution, which promises deep insights into the Linux kernel without impacting performance or requiring changes to the kernel itself. This integration with native Kubernetes environments underscores the shift towards more sophisticated and granular approaches to cloud-native observability. eBPF's capabilities enable more precise monitoring and troubleshooting, increasing both the efficiency and reliability of cloud-native systems.

KubeCon + CloudNativeCon Europe 2024 has clearly set the stage for a year of transformative developments in the cloud-native space, signaling a shift towards more sustainable, observable and edge-focused computing solutions. This convergence of innovation and technology underlines the important role of the EMULATE project in addressing the challenges and opportunities at the edge.