SLATE: Monitoring Distributed Kubernetes Clusters

Abstract

The SLATE (Services Layer at the Edge) accelerates collaborative scientific computing through a secure container orchestration framework focused on the Science DMZ, enabling creation of advanced multi-institution platforms and novel science gateways. The goal of the SLATE project is to provide a secure federation platform to simplify deployment and operation of complex and often specialized services required by multi-institution scientific collaborations, utilizing where applicable open source, cloud native tooling such as Kubernetes. This paper outlines the design and operation of a monitoring infrastructure suitable for application developers and resource providers which gives visibility to resource utilization and service deployments across a network of independently managed Kubernetes clusters.