Datadog

Connecting Datadog lets Traversal use your logs, metrics, traces, Real User Monitoring (RUM), and monitors during investigations so it can surface anomalies, correlate signals, and generate grounded root-cause explanations.

​

What Traversal reads

• Logs — queried from your log indexes
• Metrics — time-series data and tag metadata
• Spans / Traces — APM span data
• Real User Monitoring — sessions, views, actions, errors, resources, and long tasks
• Incidents — active and historical incident records
• Events — event stream data
• Monitors — monitor state and alert history
• Dashboards — dashboard panel definitions

​

Setup

​

Use Datadog during investigations

​

Logs

Use logs to find errors and operational messages, compare patterns across services or environments, and inspect the context around a failure. Example prompts:

• Find error and critical logs from the checkout service during the incident.
• Compare database timeout logs before and after the deployment.

​

Metrics

Use metrics to identify anomalous time series, correlate infrastructure and application behavior, and verify whether a mitigation restored normal behavior. Example prompts:

• Graph request latency, error rate, and throughput for checkout during the incident.
• Which Kubernetes resource metrics changed most when the failures began?

​

Spans and traces

Use traces to find slow or failed requests, identify anomalous spans, follow a request across services, or inspect a specific trace ID. Example prompts:

• Find an anomalous error trace for the checkout service during the incident.
• Inspect trace 1234567890 and identify the slowest or failing span.

​

Real User Monitoring

Use RUM to measure affected sessions or users, isolate frontend regressions by route or cohort, inspect browser errors and failed resources, and compare page performance across releases. RUM is available to organizations with RUM access enabled in Traversal. Example prompts:

• Which routes had the most affected sessions? Exclude synthetic traffic.
• Compare checkout errors by application version and browser, then inspect examples from the worst cohort.

​

Incidents

Use incidents to review active and historical incident records, understand their status and severity, and connect prior response context to the current symptom. Example prompts:

• Show Datadog incidents related to checkout failures in the last 30 days.
• Summarize the timeline and current status of the active payments incident.

​

Events

Use events to find monitor firings, deployments, configuration changes, feature flag changes, and Kubernetes events around an incident. Example prompts:

• Which high-priority monitors fired when the incident began?
• Find deployments and configuration changes for checkout in the hour before the errors started.

​

Monitors

Use monitors to inspect alert definitions, queries, thresholds, options, and runbook messages, or to find monitors by state, service, team, or tag. Example prompts:

• Show the query and thresholds for Datadog monitor 12345.
• Find alerting monitors for checkout in production and summarize their current states.

​

Dashboards

Use dashboards to discover the signals your team already watches, inspect panel queries, and use those queries as starting points for deeper analysis. Example prompts:

• Find the checkout overview dashboard and explain which panels changed during the incident.
• Use the latency panel query from the payments dashboard to compare the affected and healthy periods.

​

More information

• Datadog API and Application keys
• Datadog OAuth scopes
• Datadog site URLs