2020-03-28, 10:25–10:55, Room 1
So you want to know how your applications perform at the kernel level? Combining the power of eBPF with OpenTelemetry provides system operators with the tools necessary to collect, aggregate and visualize telemetry information about systems at large scale.
The extended Berkeley Packet Filter (eBPF) introduced in kernel 3.15 allows processes in user space to attach to tracepoints in the kernel to provide better visibility into what systems are actually doing. Much progress has been made since its introduction, and today eBPF is used in a wide range of scenarios from mitigating distributed denial of service attacks to generating flamegraphs about applications running on a host.
This talk walks through setting up tracepoints using eBPF, collecting and aggregating the data using OpenTelemetry and using the collected data to identify the root cause of a problem.