CNCF-hosted Co-located Events Europe 2025

Edge computing with Kubernetes is powerful but comes with challenges like limited resources, unreliable connectivity, and ransomware threats. Keeping systems running at the edge is critical, but how do you make that happen?

In this session, we'll share how a global retail chain used Kubernetes-native tools to recover from a ransomware attack. With 500+ stores relying on Kubernetes for inventory and sales they couldn't afford any downtime. Using immutable backups and S3-compatible storage, they fully recovered in under 10 minutes, avoiding major disruption.

Learn how to build secure and resilient Kubernetes environments at the edge, ensuring your systems are prepared for the unexpected.

Many enterprises, such as new energy vehicle and smart campus, are deploying cloudnative workloads to edge nodes. However, these nodes often exist in highly distributed and high-latency network environments, bringing challenges to manage them.

However, as some maintainers of K8s have also pointed out, there're challenges in distributed and high-latency network with K8s, including: 1) the impact of the List-watch mechanism on server-side due to data retransmission(re-list frequently) in scenarios with frequent node online/offline 2) how to recover apps when nodes go offline and restart.

This topic will introduce how KubeEdge optimizes the K8s List-watch in unstable network, avoiding the impact on the server-side bandwidth by re-list frequently. Notably, users can in the data center manage workloads on massively discrete edges without awareness. Moreover, real-world user cases such as those in new energy vehicles and smart campuses will be presented to demonstrate the effect and value.

At Focke & Co, our packaging machines are a critical part of operations in factories around the world. Downtime is serious. There’s also a necessity to process information in locations without any K8s expertise. It’s dictated by the cost of sending traffic for analysis to the cloud and by data governance. Combined with the mandatory EU resilience act, K8s on the edge becomes almost required to stay compliant.

Like many manufacturers, we found it costly to send field engineers to provide support at customer sites. So we set out to connect our machinery for remote monitoring and servicing, looking to build an IoT platform that could handle scale — and be easy to deploy in the factory.

We’ll explore how we reduced time to edge K8s deployment from two days to 15 minutes and achieved centralized remote management for a fleet of various 20k+ packaging machines. We’ll share a success story of the last two years, from our pilot project to deploying devices in production.

Being in a Nigerian university, depending on what state you school sometimes means that your education falls short in terms of the problems we are exposed to and the topics we explore in class.
To fill this gap we formed NARSDA, a small group primarily focused on astronomy based research, this semester we decided to take our heads out of the cloud and focus on some local problems facing your school.
In this session we will discuss how we built an edge device to monitor atmospheric conditions in our school, how we leveraged Prometheus, grafana and MQTT to collect and generate meaningful insights from the data and finally how we plan on using the data to cut down emissions and possibly save an endangered bird native to our local.

Organizations face a critical challenge: data is growing exponentially across distributed locations, but traditional centralized processing approaches are becoming unsustainable.

Specifically, the challenges of massive data transfer costs, regulatory compliance issues, and network reliability problems are slowing the adoption of ML infrastructure. This is particularly acute in scenarios like processing data from distributed deployments or analyzing real-time sensor edge data.

This talk demonstrates implementing compute over data with Kubernetes - bringing ML compute to where data is being created. Using real-world examples, including an energy company managing 15,000 microgrids, we demonstrate processing data in place, ensuring regulatory compliance, reducing costs, and improving reliability while maintaining centralized control. We'll showcase implementation using open-source tools like Bacalhau, providing attendees with practical patterns for modernizing their ML infrastructure.