04.10.2025 –, 107
Jazyk: English
Kubernetes networking is complex, and bottlenecks like DNS failures, packet drops, or service misrouting are often hidden from traditional monitoring tools. These subtle issues can cause latency spikes and degraded application performance that are difficult to diagnose with logs or standard metrics alone.
This talk introduces how eBPF-based observability can provide deep, low-overhead visibility into Kubernetes network flows, combined with zero-code instrumentation for applications. Using real-world examples, you’ll learn how to leverage Beyla’s detailed insights, such as DNS query tracking and pod-to-pod traffic analysis, to quickly identify and resolve networking bottlenecks.
Beginners
An ex-rocket science student turned SRE and now a Developer Experience Engineer at Grafana Labs, I am a core maintainer for the Grafana Ansible Collection and the Grafana Operator, with over 2 million downloads combined.