eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
Google has rolled out a series of updates that it says will make it easier for enterprises to set up and efficiently run container workloads on the company’s cloud platform.
Among them are new automated node management capabilities, general availability of Google’s Container Optimized OS and the release of Kubernetes 1.6, the latest version of the company’s container cluster management technology.
The automated node management features are designed to help organizations ensure their container nodes are in good health and in the most current state, Google software engineer Maisem Ali, said in a blog. Up to now, administrators were responsible for ensuring the overall health of their container nodes and making sure the nodes were up-to-date, Ali said.
A new Node Auto-Upgrade feature released this week eliminates this requirement. Instead of having to manually upgrade nodes in a container cluster, administrators can set their nodes to be automatically upgraded to the latest, fully tested node releases as they become available. Once the automatic upgrade feature has been enabled, nodes in need of updates will be automatically shut down and their workloads transferred to a new node without the need for any manual action, Ali said.
Also released this week is a new Node Auto-Repair feature that organizations can use to ensure that nodes with performance issues such as crashing binaries and disk drive capacity shortages are automatically repaired. Once the capability is enabled, Google Container Engine will monitor node health status on an ongoing basis and will automatically trigger reconstruction of a faulty node if certain conditions are met.
As part of the upgrades to the node management capabilities Google this week also introduced a new feature that lets administrators cancel or roll back a node upgrade in certain situations.
According to Ali, Google had to make substantial architectural improvements to its Container Engine logic to support the new node auto-upgrade and auto-repair capabilities. “We redesigned our entire upgrade logic with an emphasis on making upgrades as non-disruptive as possible,” Ali noted.
Starting this week, Google’s Container Optimized OS, previously known as Container VM-Image, is generally available to the company’s cloud services customers. Organizations can use Container-Optimized OS to run Docker containers on Google’s Cloud Platform.
The operating system is based on the open source Chromium OS and comes pre-installed with Docker’s container runtime and the Kubernetes cluster management and orchestration technology said Saied Kazemi, a software engineer at Google on the company’s cloud platform blog.
“The key benefit of Container-Optimized OS is that we control the build, test and release cycles, providing GCP customers—including Google’s own services—enhanced kernel features and managed updates,” Kazemi noted.
Starting this week Google has also started making Kubernetes 1.6 available to its cloud customers. The new version features several updates focused on scalability, scheduling and security, according to product manager David Aronchick.
One of the most significant updates is to the number of nodes that can be managed using Kubernetes. With Kubernetes 1.6 Container Engine now can support cluster sizes of up to 5,000 nodes or about 2.5 times the 2,000 nodes that was supported previously. The increased scalability will allow organizations to run some of their largest workloads on Container Engine, Aronchick said.