Last year Microsoft and Red Hat announced Kubernetes Event-driven Autoscaling (KEDA) – a way to bring event scale for any container or workload deployed into any Kubernetes cluster. Since then, we have been blown away by the response from the community in helping to make KEDA even better.
Our mission from the beginning was to bring the additional value and benefits of serverless workloads to everyone – in an open and inclusive way. As an important milestone in this journey, we are thrilled to share that KEDA has been accepted into the Cloud Native Computing Foundation (CNCF) as a sandbox project.
By default, Kubernetes scales containers based on resource metrics like CPU and memory. This isn’t an ideal fit for event-driven and serverless workloads where you need to rapidly scale based on events – like messages on a Kafka stream or queue messages. KEDA brings this capability for any container workload, enabling you to enhance Kubernetes and automatically drive scale up and down proactively, and based directly on the rate of events as they arrive.
For us, the biggest highlight of the last year in KEDA has been the involvement of and contributions from the community. We’ve seen community members like Astronomer.io, who contributed new scalers to drive scale based on SQL events that enabled new capabilities for Apache AirFlow. We’ve also seen users like PureFacts leverage KEDA to enable executing and scaling Azure Functions in Azure Kubernetes Service. As we move forward with KEDA within the CNCF, we look forward to a continued partnership with our community.
More details about the announcement and what’s coming next can be found on the KEDA site. If you are interested in learning more about KEDA, be sure to check out our session recording from KubeCon San Diego. You can also get involved in our Slack channel or during our bi-weekly community standups.