With KubeCon and CloudNativeCon in Copenhagen this week, we saw several container-related news announcements--from management to deployment and adoption to monitoring. As more and more organizations look to adopt containers for greater development efficiency and faster application deployment as part of a larger DevOps or IT modernization initiative, the hype and tools surrounding the ecosystem continue to grow. This week was certainly no exception.
Kubernetes Container Management
Red Hat and Kubernetes announced this week a new open source toolkit for managing Kubernetes native apps. Called Operator Framework, the teams believe that it “represents the next big step for Kubernetes by using a baseline of leading practices to help lower the application development barrier on Kubernetes.”
It does so through three main components, an SDK that provides tools to build, test and package Operators (aka Kubernetes apps) without advanced understanding of the Kubernetes API; Lifecycle Manager that facilitates management of Operators on a Kubernetes cluster; and Operator Metering that meters app usage. Red Hat is not alone in looking to help make the container ecosystem easier to navigate.
Kubernetes also announced Kubeflow 0.1, a minimal set of packages to begin developing, training and deploying ML.
Kubernetes Cluster Management
As Kubernetes becomes ubiquitous, Rancher, which is touted as, “the only product that can manage all Kubernetes clusters on all clouds and implemented by any distros,” is getting greater interest and traction with our customers. This week Rancher 2.0 was announced, featuring a Rancher Kubernetes Engine, Unified Cluster Management and Application Workload Management. The open-source, enterprise Kubernetes platform for running containers in production seeks to accelerate Kubernetes adoption and increase central visibility and control.
Google, originators of Kubernetes, introduced a new Kubernetes monitoring service called Stackdriver. Designed to help developers inspect clusters, services, workloads, pods, and containers, Stackdriver has a goal to help development better understand app behavior. And, they announced gVisor at the conference, a sandbox container runtime that provides secure isolation for containers.
The AWS consulting team at Flux7 kicked off a DevOps blog series on DevOps adoption, starting with a discussion of what motivates organizations to begin an IT modernization project. The findings: most are driven by external business pressures. Subscribe to our blog to make sure you get the whole series.
And we conclude with two new AWS features that our AWS Consulting engineers were excited to see this week:
- AWS CodeBuild now allows operators the ability to test and debug their AWS CodeBuild builds with the new CodeBuild local agent. This will save operators time and make troubleshooting easier because you can simulate a CodeBuild environment locally to test commands whereas before you’d need to fully configure and run CodeBuild to test your build. You can also now locally build your apps before committing changes to the cloud.
- This week also brought news that you can now chain the launch of multiple products in AWS Service Catalog. As an AWS Service Delivery Partner for AWS Service Catalog, the technology is core to our work to ensuring consistency, governance and time to market through the provisioning of cloud resources on AWS. Now AWS Service Catalog will allow operators to chain together the launch of multiple products, giving them the flexibility to compose products with smaller AWS CloudFormation templates as building blocks rather than creating a monolithic application stack template.Thus providing independently maintained and versioned products and simple orchestration of a combined product launch.
For ongoing DevOps news and analysis, please subscribe to our weekly DevOps blog: