Google has announced that the Kubernetes version 1.9 will now be available on the Google Kubernetes Engine (GKE) as a part of the early access program, bringing better support for stateful and stateless applications as well as hardware accelerator support for machine learning workloads.
According to Google’s blog post, this release is a big milestone which makes it easy to run a wide range of application on Kubernetes without any problems regarding the underlying architecture. By adding the latest version of the Kubernetes release to its Kubernetes Engine, which provides managing, scaling, upgrading, backing up and security of container clusters.
According to Google, the Kubernetes version 1.9 update provides “greater support for stateful and stateless applications, hardware accelerator support for machine learning workloads and storage enhancements”.
With the new update, the core Workloads API, like the DaemonSet, Deployment, ReplicaSet, and StatefulSet, which run stateful and stateless workloads in Kubernetes, have moved to general availability, ensuring support and backward compatibility.
In addition, the new version also brings hardware accelerator enhancements with the Google Cloud Platform (GCP) acting like a great environment for machine learning and data analytics workloads in containers. The new update brings improved support for Nvidia Tesla P100 and K80 GPU accelerators, aimed at anything from “genomics and computational finance to recommendation systems and simulations.
The new update also brings several storage enhancements including improvements to the Kubernetes scheduler, which simplifies the use of local storage in Kubernetes. The local persistent storage feature (alpha) enables easy access to local SSD on the Google Cloud Platform (GCP) via the Kubernetes Persistent Volume Claim.
The new Kubernetes version 1.9 update on the Google Kubernetes Engine (GKE) has also brought an alpha implementation of the Container Storage Interface (CSI), allowing the use of CSI to add different storage volume plugins in Kubernetes without requiring changes to the core codebase.
Google was quite keen to implement the Kubernetes version 1.9 update which has been released last month. With major players like AWS, Microsoft and, of course, Google, supporting Kubernetes, it is obvious that other container management platforms like Docker Swarm, Mesos, CAPS, and OpenStack, are losing the container orchestration battle.