On the Trace Analysis page of an application in the Tracing Analysis console, you can view the analysis charts of different monitoring metrics for traces, the trace list, the application topology, and the table of aggregated traces.
Go to the Trace Analysis page
Enable the real-time full data analysis feature
After you specify a sampling ratio for span data, you can enable the real-time full data analysis feature to analyze the complete sampled data in real time. The analysis results show the detailed information about the relevant traces. For more information about how to specify a sampling ratio, see Span data sampling and storage.
Filter traces
View the analysis charts
The Time change trend section displays three analysis charts. The charts show the trends in the total number of spans and number of exceptional spans, the trend in the duration of spans, and the duration distribution of spans.

- Move the cursor over the statistics chart to view the statistics.
- Select a period of time to view the statistics for the specified period.
- Click legend to hide or show the data.
View spans and servers in lists
The Span name and Machine/IP lists display all the spans and servers within your Alibaba Cloud account.

- In the search box in a list, enter a keyword and click the search icon to search for spans or servers.
- Click Span quantity or Response time in the upper-right corner of a list to sort the items in the list in descending order based on the number of spans or duration.
- Click the name of a span or the IP address of a server to use the name or IP address as a filter condition.
Group traces by field
View information on the Traces tab
The Traces tab displays the ID, start time, span name, duration, and server IP address of each trace.

- Click the ID of a trace to view the waterfall chart of the trace. For more information, see View the waterfall chart of a trace.
- Move the pointer over the name of a span to view the information about the span, including the duration, start time, tags, and log events.
View the application topology
On the Application Topology tab, you can view the application topology that displays the dependencies between the current application and other applications. You can also view the request percentage, call multiplier, and duration ratio between applications. For the best performance, the application topology feature can aggregate the request data of up to 5,000 traces for the current application.

- The request percentage is calculated by dividing the number of requests involved when Application A calls Application B by the total number of requests received by Application A. For example, 100 requests are received by Application A, but only 90 requests are involved when Application A calls Application B. In this case, the request percentage of A to B is 90%. Some requests are sent to Application B because Application A may filter all the requests based on an IF statement.
- The call multiplier is calculated by dividing the number of spans involved when Application A calls Application B by the total number of spans received by Application A. For example, 100 spans are received by Application A, and 300 spans are involved when Application A calls Application B. In this case, the call multiplier of A to B is three. If the dependency from A to B is marked as 90%/3x, 90% of the requests in Application A are sent to call Application B, and Application A calls Application B three times on average.
View information on the End-to-End Aggregation tab
The End-to-End Aggregation tab aggregates traces based on span names and application names. For the best performance, the end-to-end aggregation feature can aggregate the request data of up to 5,000 traces for the current application.

- Request Count / Request Percentage: The request percentage is the ratio of the request count of the current span to the total request count. For example, the total request count is 100, and the request percentage is 10%. In this case, 10 requests are sent to call the current span. The request percentage is calculated by dividing the request count of the current span by the total request count.
- Span Amount / Request Multiplier: The request multiplier indicates how many times each request calls the current span on average. For example, 1.5x indicates that each request calls the current span 1.5 times on average. The request multiplier is calculated by dividing the span count by the request count of spans.
- Average Self Elapsed Time / Percentage: The average exclusive duration of a span excludes the average duration of its child spans. For example, Span B is a child span of Span A. If the duration of Span A is 10 ms and that of Span B is 8 ms, the exclusive duration of Span A is 2 ms. The exclusive duration of a span is calculated by deducting the sum of the duration of its child spans from the duration of the span. For asynchronous calls, the duration of child spans is not deducted.
- Exception Count / Exception Percentage: The exception percentage is the ratio of the count of exceptional requests to the total request count. For example, 3% indicates that 3% of all requests are exceptional. The exception percentage is calculated by dividing the number of exceptional requests by the total request count. The number of exceptional requests does not equal the exception count. When the request multiplier is greater than 1, an exceptional request may have multiple exceptions.
- Click the ID of a trace to view the waterfall chart of the trace. For more information, see View the waterfall chart of a trace.
- Move the pointer over the name of a span in blue to view the traces that are related to the span. Click the ID of a trace to view the waterfall chart of the trace. For more information, see View the waterfall chart of a trace.
- Click the name of an application to obtain an overview of the application. For more information, see View the key performance metrics and topology of an application.
View the waterfall chart of a trace
The Traces tab displays the span name, timeline chart, application name, start time, server IP address, and status of each trace.

