Pain Points Persist as Reliance on Kubernetes Rises

A survey of more than 1,097 IT professionals published this week finds 42% are running at least half of their business-critical applications on the Kubernetes platform, with roughly a quarter of respondents reporting they are running more than 75% of their business-critical applications on Kubernetes.

In all, 70% of respondents employing are running Kubernetes, with a total of 65% of them employing a pure vanilla instance compared to 49% using Docker Swarm and 22% that are employing Red Hat OpenShift.

Conducted by Traefik Labs, a provider of application networking software for microservices, the survey also finds 60% of respondents are running two or more Kubernetes clusters. The two most widely deployed platforms for running Kubernetes clusters are Amazon Web Services (AWS) and on-premises IT environments, virtually tied at 47% each. A total of 44% of respondents also note cost management causes friction as they scale and expand their Kubernetes clusters. Other issues include securing applications and integrating third-party tracing systems tied at 33% each.

Not surprisingly, the top pain point cited by respondents (43%) is setup and configuration, followed by maintenance and having too many platforms at 28% each. Half of respondents (50%) also said they still log in directly into their clusters to review logs manually when issues arise.

The most widely employed management tools are Grafana and Prometheus at 74% and 68%, respectively.

Finally, 60% of respondents note they are also employing multiple ingress controllers.

Marie Ponseel, vice president of marketing for Traefik Labs, says the survey makes it clear that many of the tools being employed to manage Kubernetes are not especially efficient. As Kubernetes environments become more distributed, the need for better approaches to observability will be required, she adds.

Of course, most IT teams are still trying to determine what observability means. While observability may be a core tenet of DevOps best practices, in reality, most IT organizations are relying on a mix of monitoring tools that can surface conflicting data points. In theory, observability platforms will provide more context by synthesizing logs, metrics and traces gathered from applications and infrastructure in a more unified way. That approach would simultaneously reduce the intimidation factor many IT administrators feel when they are first exposed to a Kubernetes platform. Kubernetes is simultaneously one of the most powerful and complex IT platforms to ever hit the enterprise.

Despite the increased reliance on Kubernetes to drive deployments of next-generation microservices-based applications, the long-term success of Kubernetes in the enterprise is far from assured. IT organizations could still decide the pain is not worth the gain, especially as they continue to struggle with microservices-based applications that are every bit as complex to manage as the underlying Kubernetes platform upon which they depend.

Nevertheless, the ability to dynamically scale resources up and down on demand remains the most compelling attribute of the orchestration platform. The challenge is finding a way to tap those capabilities without burning out everyone in the IT organization in the process.

Mike Vizard

Mike Vizard is a seasoned IT journalist with over 25 years of experience. He also contributed to IT Business Edge, Channel Insider, Baseline and a variety of other IT titles. Previously, Vizard was the editorial director for Ziff-Davis Enterprise as well as Editor-in-Chief for CRN and InfoWorld.

Mike Vizard has 1614 posts and counting. See all posts by Mike Vizard