Processing data in the field and edge is beneficial for many use cases. like the collection and analysis of sensor values as it is required for many AIoT applications. Leaving the controlled environment of data centers adds various and recurring challenges for operating the computing devices. With this talk, we show how we manage the execution of applications on the edge by using and deploying Eclipse IoT Packages for and on the edge.
To interact with the devices, it is sometimes desired to deploy the respective backend not only in the cloud but also to other locations like in the field or in the transport network. Such use cases often occur in domains like industrial, residential, or others. Since Eclipse IoT Packages provides Helm Charts for Kubernetes deployments you need a Kubernetes instance in such locations.
Besides the management of the devices, one cannot forget the software and operating system running on the devices when looking at the overall architecture. The increasing complexity of use cases leads to enhanced capabilities in these devices leading to expanding adoption of Linux. This amplifies the questions of how to easily set up and update the devices over their full life cycle while they operate in an inter-connected environment.
With this talk, we share our findings and learnings of operating edge devices and Eclipse IoT Packages in and for the edge.
The talk is intended for other open-source communities and more specifically technology teams that face challenges like scalable and heterogeneous device connectivity, abstracting the access to the data available at the devices or managing the software when operating and building IoT and edge scenarios.