The Kubernetes market is on a roll. In just a few years, open-source Kubernetes has grown so much in importance that we are already seeing major market consolidation and new products from marquee technology vendors. Last week, SUSE acquired Rancher. VMware launched Tanzu Kubernetes Grid earlier this year. IBM acquired Red Hat and OpenShift. Google Anthos is moving aggressively to provide hybrid/on-premises Kubernetes solutions.
Enterprise Challenges Remain Unsolved
Meanwhile, enterprises adopting cloud-native technologies continue to struggle with these challenges:
- Talent scarcity in enterprise IT/Ops to reliably operate Kubernetes at scale with robust SLA’s
- VM’s will need to co-exist and work with containers for a long time to come; there are no easy open-source solutions that help to manage them together.
- Lock-in continues to be a concern whether the topic is public cloud infrastructure, operating systems, virtualization stacks, or hardware.
Managing thousands of Kubernetes clusters in edge locations remains an unsolved problem.
Enterprises typically choose one or more of these approaches for Kubernetes:
None of these approaches solve all of the challenges outlined above. Furthermore, using one or more of these approaches, enterprises end up with a variety of tools and technologies leading to sprawl, low utilization, and high costs.
SaaS Management plane: Ideal architecture for managing Kubernetes
At Platform9, we have pioneered a SaaS management plane architecture that provides the operational simplicity and ease-of-use of public clouds while simultaneously delivering the most open environment using upstream Open Source Stacks, e.g., Kubernetes, OpenStack, Prometheus.
The recent Gartner report “The Many Faces of Private Cloud”¹ provides information about “as-a-service” (XaaS) implementation models. Regarding the outsourced private cloud model, the report states that
“These solutions are based on an outsourced control plane (usually VM IaaS or Kubernetes) that is run as a SaaS-style offering by a provider. In this case, however, the physical hardware remains on the customer’s premises and is managed through agents or gateways from the SaaS control plane. This type of environment allows the customer to maintain hardware in a private location for compliance or regulatory reasons but outsources the complexity of running the control plane to the provider. It can also be used to support edge use cases when the customer does not want to manage the control plane.”
We have been offering such a SaaS-style control plane commercially for our Managed Kubernetes solution since 2017 and we believe this is the right architecture for enterprise adoption of Kubernetes, whether that’s on-premises, public clouds, or the emerging edge.
Central management of Edge Clouds
As edge computing takes off, management of large numbers of edge clusters centrally with a single pane of glass visibility and robust SLA’s becomes increasingly critical. The geographic distribution of edge infrastructure and workloads limits the reach of the public cloud. It can be daunting to centrally operate thousands of edge Kubernetes clusters with vendor distributions or open-source software. These clusters need to be managed centrally, with little to no “touch”. Lack of qualified engineering and operational skills adds to the challenge.
SaaS-based management provides a central operational console and a single pane of glass for all Kubernetes clusters no matter where they are located. Enterprises no longer have to worry about the operational burden of up-time/SLA management, upgrades, security patches, and production outages. This approach will be the next-generation de-facto standard for distributed cloud management whether that is VM’s or containers or bare-metal, all of which will co-exist for a long time to come.
Closing thought: Plan your enterprise strategy intelligently
Whether you have mission-critical Enterprise Kubernetes initiatives in production or you are just starting out on your containerization journey, you must be wondering how all this market consolidation is going to impact you.
As the Kubernetes market consolidates, enterprises will need to contend with their environment sprawl and its resulting operational costs and management complexity. This problem can be solved only by a centralized SaaS control plane that can abstract the differences amongst these environments and make it easy to standardize and scale operations across the infrastructure.
Enterprises would do well to closely inspect the fine print on announcements in the market and compare alternatives such as Platform9 (the only independent managed K8s provider) for operational simplicity, lock-in considerations, feature parity, as well as the track record of large scale production operations.
We have created an evaluation guide which describes the factors that lead to operational complexity and lock-in and 19 detailed capabilities to consider while evaluating an enterprise Kubernetes platform and to what extent current leading market solutions deliver these capabilities.
Kubernetes Evaluation Guide19 key consierations while evaluating an enterprise Kubernetes platform including:
- Provisioning of Kubernetes Clusters
- High Availability and Healing
- Deployment Models Supported
- And more.
- Introducing Platform9 : A better way to go cloud native (Going Cloud Native, Part 2) - August 17, 2022
- Keeping up with cloud native is hard (Going Cloud Native, Part 1) - August 17, 2022
- Grade your cloud-native maturity and find out how you can improve - December 13, 2021