Query traces in dashboards

Viewing a dashboard can provide insights such as spikes in memory usage, CPU utilization, and other metrics that help identify issues in your system. When troubleshooting problems, context matters. Knowing whether relevant trace data exists for an existing problem that's visualized in your dashboard can help decrease Mean Time To Repair (MTTR).

Rather than leave your dashboard to explore traces, you can bring trace data to where you need it and discover traces from any dashboard in your Chronosphere Observability Platform tenant.

You must first enable trace querying on each dashboard from where you want to query traces. After enabling this feature, you can query traces from a dashboard.

Enable trace querying

Before you can query traces from a dashboard, you must enable the trace query feature, which adds a Query traces button in your dashboard. After enabling the feature, you can click this button to display a dialog for querying traces directly from the dashboard.

To enable trace querying from a dashboard:

1. In the navigation menu select Dashboards.

   Alternatively, you can search for a dashboard from anywhere in Observability Platform.

2. Select the dashboard where you want to enable queries for trace data.

3. Click Settings, and then click General to display the dashboard settings.

4. In the Dashboard settings page, in the Select a tag field, type and then select tracing-context.

5. Click Apply, and then click Save in your dashboard.

The Query traces button displays at the top of your dashboard. You can now click that button to query traces.

Query traces from a dashboard

After enabling trace querying, you can display a query builder and search traces directly from a dashboard where trace querying is active.

To search traces from a dashboard:

1. In the dashboard where you enabled trace querying, click Query traces to open the Traces drawer.

2. In the Query builder, enter a service, operation, tag, or combination of those criteria to query trace data.

   For more information about how to construct a trace query, see Search and filter trace data.

3. Select a filter to apply to your query. For example, you can choose to filter the query results to only traces where the selected service or operation contains an error.

4. Click Run to run the defined query. The Query builder section collapses, and the timeline graph updates based on the query results.

   The time window displays the last 24 hours by default. You can query data up to 30 days in the past.

   If you click Run without defining any criteria in the query, results display for all traces in the last 24 hours.

5. Explore the results of your trace query. When selecting one of these options, Observability Platform applies the context from your query to Trace Explorer:

   • View trace details: To open the full trace details for an individual trace in Trace Explorer, click any trace in the table.

   • Explore queries: To open the defined query in Trace Explorer, click Open in Trace Explorer.

   • Differential diagnosis: To identify trends and scan through all related tags and values in the defined query, click Differential diagnosis. See Differential diagnosis for more information.