The Cloud Native Computing Foundation® (CNCF®), which builds sustainable ecosystems for cloud native software, today announced the findings of the fifth CNCF End User Technology Radar, a guide to a set of emerging technologies based on the experience of the CNCF End User Community. The theme of this edition for the second quarter of 2021 was Multicluster Management.
The growing popularity of distributed and multi-cloud deployments means organizations are running Kubernetes across many environments, including on-premise, public cloud, and edge. While this improves application availability, reduces latency, and makes it easier to scale both legacy and cloud native applications, it increases complexity as each cluster needs to be individually deployed and managed. The Radar team chose this as a topic because they found their organizations were all managing and deploying clusters in different ways.
“Multicluster management is still an emerging and complex space, so we wanted to gain a broader understanding of how organizations are doing it in the real world,” said Cheryl Hung, VP Ecosystem, Cloud Native Computing Foundation. “What we found is that there is no one size fits all solution for cluster management. Overall, approaches vary widely based on the number of clusters deployed and the management tools already in place, and custom tooling is usually necessary.”
After reviewing the data provided by end user organizations, the Radar team divided the solutions into two sub-categories, cluster deployment and core services and add-ons, to differentiate between tools used for deployment and those used to manage day 2 operations. The team summarized its findings in five key themes, which can be viewed in more detail on the Radar page.
Tools for managing core services, like DNS, metrics, logs, RBAC, and namespaces, outnumbered the tools for deployment on the Radar. They also received a higher total number of “Adopt” votes than deployment tools. The most widely used core services solutions are Helm, Operators, Kustomize, and GitOps tools Argo and Flux. Custom tools are used across both categories, while public and private managed Kubernetes and Terraform are popular solutions for cluster deployment. Only two solutions ended up in the “Assess” category, kOps and Cluster API, both solutions for cluster deployment. Many end users noted they are closely watching the development Cluster API, a Kubernetes sub-project, hoping it will reduce the need for custom tooling. Jsonnet for cluster services was the only tool in “Trial”.
“Compared with other Radars, we had more tools than ever in the Adopt category,” said Gabe Jackson, Cloud Platform Tech Lead at Mattermost. “This actually demonstrates how fractured the cluster management space currently is. Many organizations were almost forced into adopting many different specific tools for specific uses. As time goes on and Kubernetes deployments mature, we’ll see more consolidation of tools coming together to address a broader range of needs.
“Everyone starts with one Kubernetes cluster and expands and grows their environments from there, but currently there is no clear path for organizations starting this journey,” said Federico Hernandez, Principal Engineer at Meltwater. “In creating this Radar we wanted to document the available tools and processes so that organizations that are just starting can learn from organizations who have already been down this path.”
The CNCF Technology Radar is an initiative from the CNCF End User Community, a group of more than 140 leading-edge companies and startups, such as Airbnb, Capital One, and Twitter, who use cloud native technologies and aim to identify challenges and best practices when adopting them. The Technology Radar shares insight into which tools are used by end users and how and which tools end users recommend for broad adoption.
To learn more about the Radar results, watch the webinar with the Radar team and visit radar.cncf.io. You can also view previous Technology Radars on Continuous Delivery, Observability, Database Storage, and Secrets Management.