When an alert goes off because a Service Level Objective (SLO) is in danger of violation, it comes with a lot of context about what has been going wrong and for how long. Then Honeycomb gives you tools to explore the where & why.
Here, Martin Thwaites walks through an example of diagnosing slower performance. What service is the problem, and under what circumstances?

00:00 - Start

00:12 - What are SLOs

01:16 - SLO Burn Alerts

01:31 - SLOs and BubbleUp Anomaly Detection

02:11 - SLOs and Heatmaps

02:49 - Investigating with a Distributed Tracing Waterfall

03:41 - Heatmaps and BubbleUp

04:56 - Verifying your analysis with Trace Level queries

05:47 - Summary of SLOs and why we use them