Trace Details¶

In the Trace Explorer, click any trace to view its details, including relative time, duration, HTTP method, URL, status code, Trace ID, Flame Graph, Span list, service call relationships, resource call relationships, and related logs, hosts, metrics, and network information. For frontend applications (such as browsers), you can also view the request timing distribution, including the duration ratios of queue, first byte, and download.

When you open trace details, the system automatically collapses the quick filter panel on the left side of the Explorer to widen the detail charts and displays the Waterfall Chart or Flame Graph first according to the "Trace Details Default View" option in Preferences.

You can use the layout switcher to move extended attributes and other information to the bottom or right of the charts, adjusting the detail page layout as needed.

When you open a record from a cross-workspace query result, the details and related data are queried in the source workspace of that record. When switching to the previous or next record, the source workspace of the new record is used; a service with the same name or the same identifier cannot be used as a substitute for source determination. If the source workspace cannot be confirmed, the details entry is unavailable; refresh and try again. For query scope and permission details, see Cross-Workspace Query.

Flame Graph¶

The Flame Graph clearly shows the flow and execution time of each Span. On the right side, you can view the service list and response time. Click a Span to view its JSON content in the trace details, and use the mouse wheel to zoom for specific information.

For more details, refer to Using Flame Graphs to Analyze Trace Performance.

Span List¶

• Displays the resource name, count, duration, execution time, and ratio of all Spans in the trace.
• Enter a resource name or Span ID to search for matches.
• Click any Span to view its JSON content in the trace details; the Flame Graph updates synchronously to highlight the same Span.
• If errors exist, error indicators are displayed.
• Click Error Spans to directly display the results that match the filter.

Waterfall Chart¶

You can view the parent-child relationships between resources.

The Waterfall Chart displays Span data in order of start time. The list on the left shows Span data, including the execution time ratio of each resource. The right side displays the waterfall chart in chronological order.

• Enter the resource name or Span ID of a Span to search for matches.
• Collapse or expand the resources on the left with one click.
• Click to toggle the format of the execution time.
• Click Error Spans to directly display the results that match the filter.

Service Call Relationships¶

The service call relationships display the call relationships and performance information between different services in a rectangular diagram, including execution time, average duration, call count, total requests, average request rate, P99 response time, and error count. Requests and related metrics are calculated based on entry Spans.

By default, the current trace is analyzed. After enabling the time widget, you can select a time range to analyze the service call data within that range. Requests per second are also displayed, and the execution time ratio is hidden.

Click any service node to view its details:

• Service operations and resources: Lists the specific operations (such as API endpoints, methods, etc.) included in the service and their related resources.

• Error status tracking:

  • If the service has errors, the error details are displayed directly.

  • Click it to view the parent Span of the service, which is used to trace the upstream call source.

Uninstrumented HTTP External Services¶

When a call chain passes through an uninstrumented HTTP external service, the system identifies the service based on the upstream and downstream Span information and generates a virtual Gateway node in the service call relationships.

The call relationship may be displayed as:

• When a unique caller can be identified: Caller Service → Gateway → Downstream Service;
• When a unique caller cannot be identified: Root → Gateway → Downstream Service.

All of the following conditions must be met to generate a Gateway node:

• The downstream is an HTTP Server Span;
• The Server Span's parent_id is valid and not 0;
• The corresponding parent Span is still not found after a supplementary query;
• The Span contains a recognizable external service address.

A Gateway node is not generated when only HTTP Client Spans exist.

The virtual Gateway has no standalone Span, so call count, duration, and error count are not calculated; the related metrics display as -, and the node is shown in a neutral state.

Resource Call Relationships¶

In the trace details, select a Span, then switch to the "Resource Call Relationships" tab to the right of "Service Call Relationships" to view the upstream and downstream call relationships of the service and resource to which the Span belongs.

The relationship diagram centers on the resource corresponding to the selected Span and displays the nodes and traces that have direct or indirect call relationships with it within the selected time range. The scope is not limited to the current trace; resources with the same name are distinguished by their owning service. When cross-workspace query is enabled, only workspaces that the current account has permission to access are queried.

The diagram provides an independent time widget in the upper-right corner. Adjusting the time range updates only the resource call relationship diagram and does not affect the outer list or other detail tabs. Auto-refresh is disabled by default; manual refresh re-queries the currently selected time range.

After switching Spans, the system updates the central resource and reinitializes the time range based on the new Span's start time. When switching tabs within the same Span, the adjusted time range is retained.

The diagram highlights the central resource node and provides node instance information, a legend, zoom, fit-to-canvas, and full-screen operations. When entering or exiting full-screen mode, the current resource, time range, and query results are retained.

• Multiple levels of upstream and downstream call relationships are displayed by default, with hints for nodes that can be further expanded; you can continue expanding level by level.
• Resources at the same level are sorted by response time.
• Click a call connection line to view the key metrics and trends of the corresponding call relationship.

When a Span lacks the necessary service or resource fields, the page shows the reason why the query cannot be performed. If there is no call relationship data, adjust the time range and try again. If the query fails, retry as prompted.

Quick Operations¶

Extended Attributes¶

• In the search bar, enter a field name or value to quickly search and locate.

• After selecting a field alias, you can view it after the field name; select as needed.

• On the trace details page, you can view the field attributes related to the current trace in Extended Attributes:

Figure 1

Error Details¶

On the trace details page, if the current trace contains errors, select the error span filter option in the sidebar. The system automatically highlights all error Spans. Click any highlighted Span to view its associated error stack, logs, and other details.

For more details, refer to Error Tracking.

Service Context¶

By retrieving the object categories in the Infrastructure Resource Catalog and displaying the latest object based on create_time, you can quickly view the runtime information, service dependencies, and integration information of the current service.

Related Analysis¶