Honeycomb, the observability platform that enables engineering teams to find and solve problems they couldn't before, is announcing the launch of Honeycomb for Kubernetes, expanding Honeycomb’s observability to Kubernetes workflows.
As Kubernetes becomes the preferred environment for managing containerized applications, a gap in application context between Kubernetes and infrastructure layers generates a variety of inefficiencies, including operational backlog, strained resources, and decreased productivity, according to Honeycomb.
While there are a myriad of infrastructure dashboards within application performance monitoring (APM) suites that aim to streamline the observation of notoriously complex Kubernetes workloads, they fail to deliver in-depth application context which can be crucial in diagnosing and resolving software issues.
“When we surveyed the landscape, we saw that there's a lot of the old school monitoring companies that are doing dashboards, Kubernetes, etc., from a very infrastructure-first perspective, and then you've got a new generation of Kubernetes tooling that is sort of standalone Kubernetes for understanding and monitoring it. And the problem is that these data sources aren't really linked,” explained Charity Majors, CTO of Honeycomb.
“You've got most companies using…one [tool] for metrics, and they buy another tool, they use it for logs, they use another tool for traces, use another tool for security, they use another tool for Kubernetes. Every single time you add a new tool, you have to pay to store that data again, and then it becomes increasingly lossy and difficult for the engineer who's sitting in the middle,” she continued.
Honeycomb for Kubernetes is engineered to reunite observability and developers working on Kubernetes, enabling enterprise teams to better differentiate application and infrastructure issues. By addressing potential bottlenecks between platform engineers and developers, overall, these teams are afforded greater confidence in their releases, more seamless migrations, higher self-sufficiency, and improved productivity.
The solution’s technology correlates application requests with particular Kubernetes pods, nodes, or cluster configurations, simplifying data integration from Kubernetes with new instrumentation options for Kubernetes events, metrics, and trace attributes. The Honeycomb UI further streamlines observability, allowing users to easily connect Kubernetes context to any incident’s events and uncover relevant patterns.
“The thing that we did for Kubernetes [that] we've actually never really done before was really looking at the workflow from the engineer’s perspective and…making it so that it's intuitive, it's easy to set up,” said Majors. “Whether you have existing code or whether you're just starting up from scratch, we pick the right defaults, we pre-generate super helpful dashboards…we feel like Kubernetes is the new Linux, so it made sense to make that an elegant, beautiful experience from end-to-end.”
Furthermore, by narrowing down to a single source of truth with Honeycomb, enterprises pay for storage once and receive the informational benefits of Honeycomb’s observability. On top of that, spend is predictable; adding attributes is zero cost—increasing dimensions and enriching data context does not create additional spend.
Honeycomb for Kubernetes is a scalable, customizable solution that features exponential telemetry from pods to nodes and unlimited custom attribute tags per event at no extra cost. Additionally, human-driven, AI-assisted investigation via Honeycomb's Query Assistant enables developers to use natural language to query and understand aspects of application performance associated with Kubernetes.
“Anytime you're dealing with these really complex systems, it's hard to know where to start,” said Majors. “Part of our philosophy is that we want to let computers do what computers are best at so that we can help humans do what humans are best at…Query Assistant is, I think, a beautiful example of that. It lets humans do what humans do best—which is interpret the data, understand it, attach meaning to things, follow the trail of breadcrumbs to something that is meaningful, while letting computers really do the heavy lifting where the numbers are concerned.”
Addressing data and tool fragmentation, Honeycomb is introducing support for OpenTelemetry Kubernetes standards, which streamlines the instrumentation of cluster context into application traces. Ultimately, this enhances open source instrumentation on behalf of platform teams, where the frequency of code instrumentation is reduced and therefore increases overall efficiency.
Honeycomb for Kubernetes is compatible with Amazon EKS, Azure Kubernetes Service, and Google Kubernetes Engine, as well as bare-metal Kubernetes distributions and Red Hat OpenShift.
To learn more about Honeycomb for Kubernetes, please visit https://www.honeycomb.io/.