Control Observability Costs With Sampling

How can we send less data without losing information? Sampling removes duplication among distributed traces, while keeping all the interesting ones. See how this works with Honeycomb and Refinery.

00:23 - One Hundred Thousand Copies of the Same Trace

00:41 - What is Refinery?

01:00 - Typical Sampling Rule Configuration Example

01:51 - Events We Were Charged for, vs Represented