SUSE Rancher and K3s running on the SUSE Linux Enterprise Micro operating system were deployed on Dell VxRail hyperconverged infrastructure as part of ongoing technology validation exercises performed by a team of SUSE and Dell VxRail engineers. This joint effort was a continuation of activities resulting from the partnership between Dell Technologies and SUSE that seeks to bring validated modern products and solutions to market that enable our customers to operate CNCF-Certified Kubernetes clusters from the core to the cloud and to the edge, to support their digital businesses and harness the power of their data.
SUSE Rancher is a complete software stack for teams that are adopting containers. It addresses the operational and security challenges of managing multiple Kubernetes (K8s) clusters, including lightweight K3s clusters, across any infrastructure, while providing DevOps teams with integrated tools for running containerized workloads. It not only deploys production-grade Kubernetes clusters from the datacenter, to the cloud, to the edge, but it also unites them with centralized authentication, access control, and observability. Included as a GitOps feature of Rancher is Fleet, a container management and deployment engine that is designed to manage up to a million clusters. Rancher is available as a free download from the SUSE Rancher GitHub.
K3s is a CNCF sandbox project that delivers a lightweight yet powerful certified Kubernetes distribution. When used with SUSE Rancher, K3s is ideal for running production workloads across resource-restrained, remote locations, or on IoT devices. K3s is packaged as a single < 40 MB binary that reduces the dependencies and steps needed to install, run, and auto-update a production Kubernetes cluster.
SUSE Linux Enterprise Micro (SLE Micro) is an ultra-reliable, lightweight operating system purpose built for containerized and virtualized workloads. It leverages the enterprise-hardened security and compliance components of SUSE Linux Enterprise and merges them with a modern, immutable, developer-friendly OS platform, making it a great fit for running K3s clusters.
VxRail is the only fully integrated, pre-configured, and tested HCI system optimized with VMware vSphere, making it ideal for customers who want to leverage SUSE Rancher, K3s, and SLE Micro through vSphere to create and operate lightweight Kubernetes clusters on-premises or at the edge. Running K3s along with SUSE Rancher on VxRail delivers a seamless and automated operations experience across cloud-native and traditional workloads.
VxRail’s intelligent lifecycle management automates non-disruptive upgrades and patches while keeping the infrastructure in a continuously validated state to ensure workloads are running and clusters are optimized. Together, VxRail, SUSE Rancher, K3s, and SLE Micro make it easy to standardize both IT and developer operations on-premises in the core, or edge, and in the public cloud.
In this non-production lab validation, a single VxRail cluster was deployed with the following specifications:
4 x VxRail E660F nodes running VxRail 7.0.370 version software
After the VxRail cluster was built, a set of three virtual machines were deployed running SLE Micro 5.1. A multi-node K3s cluster running version 1.23.6 with Server and Agent services, Etcd, and a ContainerD container runtime were installed on these VMs. SUSE Rancher 2.6.3 was then installed on the K3s cluster. Also included in the K3s for Rancher installation were Fleet GitOps services, Prometheus monitoring and metrics capture services, and Grafana metrics visualization services. All of this formed our Rancher Management Server. Rancher was then used to deploy managed K3s workload clusters that were single node and multi-node in size. In this validation, two managed K3s workload clusters were deployed on the Rancher Management Server cluster. These included single node and six-node K3s clusters all running on vSphere VMs with the SLE Micro operating system installed. Note: All the installation steps used followed the SUSE documentation without any unique VxRail customization.
This validation can easily be modified to be more highly available and production ready. The following diagram depicts how to incorporate more resilience:
The Rancher Management Server stays the same, because it was already deployed with a highly available four-node VxRail cluster and three SLE Micro VMs running a multi-node K3s cluster. As a production best practice, managed K3s workload clusters should run on separate highly available infrastructure from the Rancher Management Server to maintain separation of management and workloads. In this case, a second four-node VxRail cluster can be deployed. For the managed K3s workload clusters, a minimum of three-node clusters should be used to provide high availability for both the Etcd services and workloads running on it. However, three nodes would not be enough to provide node separation and high availability for the Etcd services and workloads. To remedy this, customers can deploy a minimum six-node K3s cluster. This is depicted in the diagram with the K3s “Kubernetes 2 Prod” cluster.
This validation features Dell VxRail. Customers have the flexibility to deploy similar architectures using other Dell hardware platforms, such as Dell PowerEdge and Dell vSAN Ready Nodes running VMware vSphere. For more information, and to see other jointly validated reference architectures using Dell infrastructure with SUSE Rancher, K3s, and more, see the following resource pages and documentation.