Viewing and analyzing traces
The trace view page displays the end-to-end request path: from the root span to final operations. To open a trace, use search or type in its trace_id.
Gantt chart
The trace is visualized as a Gantt chart, where each horizontal bar represents a span and its length indicates the operation duration. Nested spans are positioned beneath their parents, forming a call tree. Any spans with errors are highlighted in color.
Span details
Click a span to open its details. The information is organized into tabs:
- Attributes: Key-value pairs added during instrumentation: HTTP method, URL, database table name, etc.
- Events: Timestamped data recorded within the span (e.g., an
exceptionevent). - Links: References to associated spans within the same or a different trace. They are used in asynchronous operations. If a span has links, they are displayed next to its name.
The following actions are available from the span details:
- Share: Copies a link to the selected span to your clipboard.
- Go to logs: Opens logs for with the span or entire trace. For more information, see Trace-to-log correlation.
Analysis strategies
Top-down
Use this approach when you lack prior context, e.g., when you receive a trace_id from an alert or a user. Start at the root span and follow the request’s path through each service to pinpoint latency hotspots or the exact point of failure.
Bottom-up
Use this when you are investigating a specific service or operation. Use the in-trace search bar to find relevant spans: matching spans will be highlighted.
Interface actions
- Show only found results: Hides irrelevant spans, displaying only those that match your search.
- Spans with errors: Filters the trace to show only spans with
ERRORstatus. - Spans on critical path: Displays spans affecting the total request execution time. For more information, see Critical path analysis.
- Share: Copies a link to the trace to your clipboard. If a specific span is open, the link will point directly to that span.