Lead Image © Lightwise, 123RF.com
Manage VMs and Containers with Harvester
Data Farming
Broadcom's acquisition of VMware primarily resulted in significantly higher costs for vSphere customers, which explains why many vSphere users are looking for an alternative – such as KubeVirt [1], which joins forces with OpenShift or Rancher Prime to manage traditional virtual machines (VMs) and containers. Of course, this arrangement primarily happens in large-scale installations with dozens of nodes. Regular readers of ADMIN will have already discovered how KubeVirt works [2].
SUSE's Harvester [3] steps up to show that KubeVirt is also a great choice for small to medium-sized businesses (SMBs). The software is open source and available free of charge on GitHub, and support is available for the commercial Harvester variant named SUSE Virtualization [4]. In this article, I take a close look at the tool and its features and offer tips for simplifying your work with Harvester and expanding its functionality. As a result, Harvester will run perfectly even in SMB environments.
Generous Hardware
The SUSE Linux Enterprise (SLE) Micro distribution, a heavily pared-down version of SUSE Linux, runs on a physical Harvester node. SLE Micro is image-based and immutable, meaning the root filesystem is mounted read-only, and no changes can be made to the base system; not even the Zypper package manager will work here. When a new version of the operating system (OS) is available, Harvester replaces the node's entire image. Embedded within the OS image is the next-generation Rancher Kubernetes Engine (RKE2) distribution. All other functions run in Kubernetes pods.
The hardware requirements have changed slightly since 2024. Harvester now requires at least eight CPU cores and
...
Buy this article as PDF
(incl. VAT)