production-investigation

You can debug production issues in Honeycomb using a structured workflow that guides you from initial alerts to root cause identification via context gathering, broad querying, BubbleUp analysis, and trace examination.

The best way to investigate latency spikes or error surges is following a step-by-step process: orient, characterize, use BubbleUp to find differentiators, drill into traces, verify, and record the findings.

BubbleUp analysis works for root cause analysis by finding differentiators in your query data, allowing you to identify specific causes of production issues such as a particular deployment version causing a latency spike.

Yes, you can use this workflow to debug deployment regressions and dependency failures, as it provides structured investigation workflows specifically addressing these production issue scenarios.

No dependencies are required to automate incident response with this process, as it orchestrates Honeycomb tool calls internally to investigate production issues without external dependencies.

Examining individual request flows involves using the get_trace functionality to analyze traces after identifying differentiators with BubbleUp, allowing you to confirm the specific bottleneck causing the production issue.