In the rapidly evolving Kubernetes ecosystem, groundcover is establishing itself as a disruptive force in the observability market through its innovative use of eBPF technology. In this exclusive VMblog interview, Orr Benjamin, VP of Product at groundcover, discusses how the company’s kernel-level approach to data collection provides advantages over legacy observability tools, particularly for cloud-native environments. Benjamin reveals details about groundcover’s new “1-click” Kubernetes dashboards launching at KubeCon, explains their unified data model that enables seamless transitions between high-level visualizations and granular exploration, and elaborates on the company’s distinctive “Bring Your Own Cloud” approach that combines SaaS convenience with the security benefits of on-premises data storage.
VMblog: groundcover is known as an observability company that’s fully embraced eBPF. What unique advantages does that give you in the market over legacy observability players? Especially in the Kubernetes ecosystem.
Orr Benjamin: At groundcover, we’ve embraced eBPF because of its unique ability to deliver high-performance observability across cloud-native environments. This gives us a significant edge over legacy observability tools, particularly within the Kubernetes ecosystem. eBPF allows us to collect data directly from the kernel, providing unmatched visibility into the inner workings of containers, pods, and the underlying infrastructure without requiring code changes or creating unwanted overhead. Unlike legacy tools that rely on agents or sidecars, eBPF operates efficiently at the kernel level, making it lighter and faster while providing more granular insights.
Traditional observability tools also often rely on sampling to reduce overhead, which can lead to blind spots or incomplete data. With eBPF, we capture every event and metric in real-time, ensuring that no data is missed, which is crucial for high-demand, dynamic Kubernetes environments where things change rapidly.
VMblog: You are launching “1-click” Kubernetes dashboards here at KubeCon. Can you explain how it’s possible to spin up a dashboard with relevant application / business context that quickly?
Benjamin: Groundcover’s 1-click Kubernetes dashboards are built to streamline the process of generating context-rich visualizations by leveraging eBPF for real-time, kernel-level data collection across the entire Kubernetes stack. By automatically aggregating key metrics and traces from applications, infrastructure, and system health, these dashboards can be spun up instantly with relevant insights tailored to the user’s technology stack. The integration with OpenTelemetry further enhances the ability to trace service interactions and uncover additional context like user IDs and error messages.
The single click functionality is powered by a seamless integration between groundcover’s observability platform and the Kubernetes environment. With this approach, developers can automatically generate a dashboard with all the relevant context about the application, system, and Kubernetes node health. There is no need to manually set up data sources or configure individual metrics, which significantly reduces the setup time for users.
VMblog: One interesting thing from the press release is groundcover’s shared data model that enables users to toggle between these dashboard visualizations and deeper data exploration. How does that work and what does that experience look like for the user?
Benjamin: groundcover’s shared data model powers a seamless experience where users can easily toggle between high-level dashboard visualizations and deeper data exploration. This is made possible by the integration of eBPF and OpenTelemetry, which aggregate real-time metrics and traces across the entire Kubernetes stack, all under a unified data model.
A key piece of this experience is that everything operates through query builders that share the same structure across native pages-whether it’s logs, traces, events, dashboards, or monitors. This consistency enables users to explore native pages and find the data they need, then quickly click through to the relevant dashboard or monitor with just one action. Similarly, users can click from a dashboard to create a monitor or dive straight into the underlying data with ease.
For users, this means they can start with a dashboard that provides key insights into the health and performance of their applications, then instantly drill down into granular data without needing to manually configure anything. Whether they’re investigating an alert or diving deeper into a specific service interaction, the shared data model ensures that the context is consistent across all views. This allows users to seamlessly navigate from a macro view of system health to detailed insights into metrics, traces, and logs, all while maintaining the necessary context. The consistency and ease of navigation significantly reduce friction in troubleshooting and accelerate the time to resolution, empowering teams to make more informed decisions faster.
VMblog: I wanted to ask about the “bring your own cloud” approach that groundcover really seems to be pushing. The claim is that customers get all of the benefits of SaaS in terms of user experience, while being able to also keep all of their data on prem. How is that possible and what did groundcover have to accomplish technically to make that possible?
Benjamin: groundcover’s “Bring Your Own Cloud” (BYOC) approach enables customers to leverage the ease of SaaS while maintaining full control over their data by deploying the observability platform within their own cloud infrastructure. To make this possible, groundcover had to address several key technical challenges. First, the platform was designed to be fully containerized and cloud-agnostic, ensuring seamless deployment on any cloud environment, whether on-prem or across public clouds like AWS, Azure, or GCP. This flexibility allows customers to deploy groundcover’s observability stack without needing extensive infrastructure changes.
From a technical perspective, we developed inCloud, our version of the first full-blown BYOC Observability solution, which abstracts and simplifies the process of running the platform in a self-hosted environment. The integration of eBPF-based data collection and OpenTelemetry for distributed tracing remains consistent, regardless of whether the platform is hosted in groundcover’s own cloud or the customer’s private cloud. This ensures that customers still receive full observability across their infrastructure while keeping their sensitive data on-prem. Additionally, groundcover’s use of open, standardized protocols ensures that customers can easily integrate the observability platform with their existing tools and workflows, avoiding vendor lock-in and enabling future adaptability as the organization scales.
To support BYOC, we had to build out a flexible, self-reliant architecture that can handle dynamic scaling and data management in varied environments. The platform was engineered to work with customers’ cloud-native infrastructures, such as Kubernetes and containerized environments, ensuring that observability remains effective even as workloads evolve. The data model is designed for full data sovereignty, meaning users retain complete control over where their data is stored and who has access, enabling them to comply with security and regulatory requirements while also managing costs more effectively. Ultimately, our BYOC approach strives to bring together the best of both worlds: full control over data with the power and convenience of a modern observability stack.
##
