The Kubernetes Storage Layer: Peeling the Onion Minus the Tears - Madhav Jivrajani, VMware 

Don't miss out! Join us at our next Flagship Conference: KubeCon + CloudNativeCon Europe in Paris from March 19-22, 2024. Connect with our current graduated, incubating, and sandbox projects as the community gathers to further the education and advancement of cloud native computing. Learn more at kubecon.io
The Kubernetes Storage Layer: Peeling the Onion Minus the Tears - Madhav Jivrajani, VMware
The Kuberentes storage layer to the end user is etcd, but it actually consists of 3 different layers. In order to scale Kubernetes to the limits it can today, a significant amount of work has been done and continues to happen on the storage layet. The 3 layers of abstraction are - the watch cache, the cacher and finally etcd. Understanding how API requests interact with each layer can have significant cost reductions and performance gains, and understanding the implementation of these layers can seem challenging, but it doesn't have to be! Knowing the internals of the storage layer can greatly supplement a user's existing knowledge of their workloads and can help them starting from capacity planning all the way to writing better controllers. A rough outline of the talk is as follows: - Overview of how Kubernetes processes requests - The Cacher - The watch cache - Potential CPU and Memory hotspots at each layer - Recent work done to improve reliability and scalability - Q&A