2022 in Review: Kubernetes’ Big Year
2022 was a big year for cloud-native technologies—especially for Kubernetes, which became a platform of choice for digital transformation and cloud-native workloads. As such, much of this year’s discussion revolved around Kubernetes updates and best practices.
Below, we’ll analyze a handful of key takeaways from this year’s coverage of Kubernetes. The insights below are culled from many discussions with industry experts as well as relevant reports from the Cloud Native Computing Foundation (CNCF) and Kubernetes-related technology providers from throughout the year.
Going Mainstream
This year, growth around Kubernetes knew no bounds. An early 2022 report from CNCF found that 96% of respondents are now either using or evaluating Kubernetes. And a full 79% of respondents use managed services, like EKS, AKS or GKE.
The popular open source utility is useful for container orchestration and service discovery, load balancing and many other application life cycle capabilities. Due to its robust capabilities, most organizations that use Kubernetes report success with the platform. “Kubernetes has become mainstream,” said D2iQ CEO Tobi Knaup. “It has become an industry standard for digital transformation and next-generation workloads.”
Kubernetes Became Reachable for Most Organizations
Kubernetes was initially viewed as something only larger enterprises could benefit from. It has a steep learning curve and requires experts to operate. But due to advances in usability and managed services offerings, in 2022 Kubernetes became more accessible for small-to-medium businesses.
“We’re definitely crossing the chasm of the early innovator stage and entering the early majority stage,” says Murli Thirumale, VP and GM of Pure’s cloud-native business unit. “Now that products have become more stable, the technology is ready to enter the mid-market.”
Edge Use Increased
This year saw rising interest in using Kubernetes at the edge and within bare metal instances. In fact, 35% of production Kubernetes users are already deploying K8s at the edge, and many more plan to do so in the next 12 months, according to a study from Spectro Cloud. A full 81% of respondents say there are compelling use cases for the edge in their industry.
Several reasons may fuel the move to the edge. But the most pressing is to run high-throughput computation, such as artificial intelligence, closer to the data. This could enable better performances and avoid the ingress/egress fees involved in transferring data to and from the cloud. Edge Kubernetes can be facilitated by several CNCF open source projects, such as KubeEdge, SuperEdge, Akri and others.
The K8s Ecosystem Relies on Open Source
We’ve witnessed open source packages fueling much of the progress around Kubernetes, promoting vendor-agnostic standards in the cloud-native ecosystem. “In general, open source technologies are now driving the next era of IT innovation,” wrote Mike Vizard. Many innovative open source technologies are not just compatible with but are packaged as Kubernetes Operators. These cloud-native tools can be easily leveraged in Kubernetes distributions to enable functionality related to CI/CD, data storage, DevOps, testing and many other areas.
Stateful Deployments Became More Accepted
Although containers are designed to be ephemeral and stateless, most applications nonetheless require some form of persistent storage. And the community has developed some workarounds to bring stateful deployments to Kubernetes. Some lessons here include using a Kubernetes-native backup architecture, having a data backup plan that leverages automation, recovering in the right order and leveraging a process that’s agnostic to database types. For example, one large health care company deploying K8s in production shared how they use Portworx to handle their “polyglot” data storage types.
Exponential Challenges Persisted
Even with the aforementioned breakthroughs, many challenges persist. In fact, 86% of respondents cited exponential challenges associated with using multiple Kubernetes environments. A separate study from D2iQ found that only 42% of K8s projects actually make it into production and that one in five developers claim Kubernetes makes them feel extremely burnt out.
One major barrier is simply getting Kubernetes up and running, which can take months. After that comes operational headaches. One such challenge is overseeing multiple K8s clusters. Another is overseeing increasing complexities while running Kubernetes simultaneously on multiple environments. Companies experiencing a talent gap may lack the necessary configuration skills to handle this, making ongoing maintenance difficult.
Governance Became Necessary to Harden Kubernetes
The move to cloud-native brings new threat vectors, such as misconfigurations and insecure default settings. And Kubernetes is not immune to these burdens. A mid-2022 study from Aqua Security found increased software supply chain attack sophistication across the board, with a significant number of attacks leveraging vulnerable Kubernetes deployments. To put things in perspective, Shadowserver estimated there are 380,000 publicly deployed Kubernetes API servers.
As more engineers utilize the platform, increased governance and role-based access control become more important to protect usage. As such, a CNCF report, The State of Cloud-Native Policy Management, found a growing use of policy-based controls in Kubernetes environments. About 50% of respondents report using policy enforcement with K8s, with OPA being a common agent to direct cloud-native policy creation and enforcement. Such controls will be necessary to avoid breaking the rule of least privilege.
Final Thoughts
2022 was a big year for Kubernetes. We saw K8s enter mainstream status and become a reachable target for many more organizations. Edge use increased, and open source became cemented as a driver of cloud-native success. Yet, usability challenges persisted and supply chain security threats reared their ugly head.
So, what lies ahead in 2023? Well, cloud-native leaders foresee 2023 being another pivotal year with additional growth and maturity in this field. One such avenue is AI—88% identify Kubernetes as the platform of choice for running AI and ML workloads within the next two years. Other predictions involve more platform engineering, increased emphasis on developer experience, and—thankfully—engineers getting up-to-speed with the complexities of Kubernetes.