Zero-instrumentation Observability Based on eBPF

How do you achieve full-stack observability without modifying a single line of code? This talk explores how eBPF is changing the way teams collect telemetry data, and why it matters for modern infrastructure monitoring.

Observability is essential for any team that needs to detect and resolve issues quickly, but traditional approaches are costly: instrumenting every application is time-consuming, resource-intensive, and difficult to scale. This session examines a more efficient path.

We dive into how eBPF enables the automatic collection of metrics, logs, and traces across containerized environments with zero code changes. You'll see how it tracks container activity at the kernel level, including network calls and filesystem operations, giving teams deep visibility into distributed systems without the usual instrumentation overhead.

The session also covers how to turn that telemetry data into actionable insights, walking through practical approaches to troubleshooting and root cause analysis using the signals eBPF surfaces.

If you're an SRE or DevOps engineer looking to reduce the operational burden of observability while gaining better visibility into your Kubernetes or containerized infrastructure, this talk offers a concrete look at what zero-instrumentation observability makes possible.