Google Heats Up Container War With Kubernetes 1.3
Firing the latest volley in the container battle, Google announced the availability of Kubernetes 1.3, the latest version of the open source container orchestration engine. This release comes within a few days of Docker, Inc. announcing Docker 1.12 which comes with built-in Swarm orchestration capabilities.
Though Kubernetes is a community-driven project backed by Linux Foundation, Google is one of the prominent contributors to the project. With features such as hybrid deployments, better scale, and the ability to run traditional workloads, Google is aiming to make Kubernetes enterprise friendly. This release addresses some of the most requested features by the early adopters of Kubernetes.
Here are a few reasons why Kubernetes score better than other orchestration engines:
Performance and Scalability
When it comes to scalability, Google knows a thing or two. It’s bringing some of the internal best practices to Kubernetes. These contributions coupled with those from the community make Kubernetes one of the most scalable container platforms.
With the latest release, Kubernetes was pushed to the limits by running it on a cluster with 2000 nodes with 60000 pods. Clusters are a collection of individual machines that are called nodes. A pod in Kubernetes is the primary unit of deployment that can be compared to a virtual machine. On an average, each pod was started within 2 seconds of deploying them. What does this mean to businesses? During an unexpected surge or spike in the traffic, you would want to launch more instances of an application. In a typical public cloud deployment, where the workload runs in virtual machines, it would take anywhere between 5 to 15 minutes for the new instance to become available. The same application, when running in containers, can scale at a rapid pace ensuring better response times.
Kubernetes 1.3 can dynamically scale clusters up and down while doubling the maximum number of nodes per cluster.
Though not every business needs that level of scale and performance, it is certainly beneficial to e-commerce and other online companies would tremendously benefit from this feature. It also helps managed service providers and DevOps teams who are measured on the response times and availability of applications.
Read the entire article at Forbes
Janakiram MSV is an analyst, advisor, and architect. Follow him on Twitter, Facebook and LinkedIn.