Timeline

When should I use this?

• You need to see what changed before production behaviour changed.
• You are comparing competing hypotheses against temporal order.
• You want a shared audit trail for handoffs between teams.

Requirements

• An active or recent investigation with connected telemetryRequired
• Incident time window that covers degradation onsetRequired
• Deploy or change signals for correlation (recommended)Optional

Steps

1. Read the incident clock

The timeline orders events by time — deployments, latency changes, saturation, dependency effects, and SLO breaches. Temporal correlation is the starting point, not final causality.

2. Follow the audit trail

Investigations progress through visible stages:

Incident → Evidence → Hypotheses → Tests → Explanation

Each confidence update maps to evidence you can inspect. Confidence changes when evidence changes — including when contradictory patterns appear.

3. Compare expected vs observed order

Hypothesis tests compare expected if true against observed event order. Temporal order that contradicts a hypothesis weakens it — even when a correlating metric exists.

Optional deep dive: Walk the checkout-latency timeline on the Demo page.

Did it work?

• Timeline covers the full incident window
• Deploy or change events appear before degradation
• Confidence updates link to visible evidence

Troubleshooting