Canonical Releases Autonomous Clustering with MicroK8s


Canonical has announced a new release of MicroK8s that includes autonomous High Availability clustering.

In the latest release of MicroK8s, Canonical has introduced autonomous high availability to the lightweight clustering tool. This addition, which is automatically enabled once three or more nodes are clustered, adds resilience for production workloads in both cloud and server deployments.

According to Alex Chalkais, Product Manager at Canonical, β€œThe autonomous HA MicroK8s delivers a zero-ops experience that is perfect for distributed micro clouds and busy administrators.”

What makes the automatic autonomous high availability possible is the Dqlite datastore. This datastore is Canonical's raft-enhanced Sqlite, which is embedded inside Kubernetes. Dqlite dramatically reduces a cluster's memory footprint and automates the maintenance of the datastore.

With this new feature in place, MicroK8s will automatically choose the best nodes to provide the datastore and, in case of a node failure, the next best node is automatically promoted. An HA MicroK8s cluster will be able to withstand the loss of any node and still be able to meet production requirements with minimal cost and oversight.

This new addition will also go a long way to harden industrial IoT applications. Industry 4.0 workloads (such as AI inference and Kafka) are a perfect fit for HA Microk8s on mission-critical control systems. This increased resilience of High Availability MicroK8s will be a boon to Kubernetes clusters that exist on edge nodes, such as remote branch offices, retail POS, cell towers, and autonomous automobiles.

For more information, read the official announcement from Canonical.


Related content

  • News for Admins
    In the news: Red Hat Enterprise Linux 8.3 Available, It's Official – Linux 5.10 is the Next LTS Kernel, System76 Unleashes World's Smallest Quad-GPU Workstation, Canonical Releases Autonomous Clustering with MicroK8s.
comments powered by Disqus