What is it about?

Distributed software applications are built from communicating services. They are often deployed in the form of containers in a Kubernetes compute cluster. Lately, it has been interesting to include computers which are not directly in the data center in such Kubernetes clusters. It then becomes difficult to optimally place the services in an application depending on the communication pattern, which may also change over time. We describe a working methodology and implementation on how Kubernetes applications can be deployed in a way that saves money under dynamically changing conditions.

Featured Image

Why is it important?

Current practice is to use manual (or semi-manual) methods to place services while our methodology is entirely automatic.

Read the Original

This page is a summary of: Cost-aware service placement and scheduling in the Edge-Cloud Continuum, ACM Transactions on Architecture and Code Optimization, January 2024, ACM (Association for Computing Machinery),
DOI: 10.1145/3640823.
You can read the full text:



The following have contributed to this page