This page shows you how to perform basic operations for a project that is using Stackdriver Trace.

Trace is automatically enabled for all applications running on App Engine standard environment. For applications running on a VM or a container, including VM instances from Compute Engine and App Engine flexible environment, and containers in Google Kubernetes Engine, see Setting up Stackdriver Trace page for details on how to instrument your application in various environments and languages.

Before you begin

  1. Sign in to your Google Account.

    If you don't already have one, sign up for a new account.

  2. Select or create a GCP project.

    Go to the project selector page

  3. Make sure that billing is enabled for your Google Cloud Platform project. Learn how to enable billing.

View the trace overview

The trace overview is the default view in Trace. This page displays latency data and summary information, such as the number of trace spans ingested in the current and previous month.

Go to Stackdriver > Trace:

Go to trace overview

Find a trace

Go to the Trace list page to view a list of recent requests to your application. The trace list allows you to browse and filter all associated traces by URI, module, version, time range, and other parameters.

If you haven't deployed an application to App Engine standard environment or deployed an application that is instrumented to generate trace data, then the Trace list page displays the message There are no matched traces.

If you have trace data, you can use the trace list to find traces whose details you want to view:

Sample Stackdriver Trace list pane.

View trace details

Click the URI of any displayed trace in the trace list. The trace details appear in the Google Cloud Platform Console.

The detail view shows a summary of details about the request, a graphical timeline that shows the root span for the end-to-end request and subspans for any RPC calls, and a detailed view of latency data collected for the spans:

Sample Stackdriver Trace details pane.

Create an analysis report

An analysis report shows you an overall view of latency for a set of requests to your application.

To create an analysis report:

  1. In the navigation panel, click Analysis reports. If you have any reports, they are listed:

    Sample analysis reports page.

  2. At the top of the analysis-reports page, click + New Report.

  3. To include all requests in the report, enter the root span name into the Request filter field. In the previous figure, the root span name is /. You can also enter filter terms in this field to select a subset of traces for the analysis report. See Analysis reports for more information.

  4. Accept the remainder of the default settings for the report. For the time span, the default is the hour previous to when the report was created.

  5. Click Submit to create the report. The new report appears in the analysis report list.

View an analysis report

To view an analysis report, in the analysis report list described above, click the name of the report:

Display of a specific analysis report.

  • To view the distribution of requests along a logarithmic axis of average latency times, click the Density distribution button.

  • To view the percentage of requests along a logarithmic axis of average latency times, click the Cumulative distribution button.

The Latency panel below the graph shows the average latency times by percentage with links to traces. To view trace details for representative samples of requests at different percentiles, click a number in the Sample Traces column.

If bottlenecks have been identified, then Trace displays a Bottlenecks panel. When displayed, this panel includes links to sample traces. By clicking these links, you can review and analyze specific traces.

What's next

Was this page helpful? Let us know how we did:

Send feedback about...

Stackdriver Trace
Need help? Visit our support page.