Copy as Markdown[Open in ChatGPT](https://chatgpt.com/?q=Read%20https%3A%2F%2Fcoralogix.com%2Fdocs%2Fuser-guides%2Fapm-v2%2Ffeatures%2Ftraces.md%20and%20help%20me%20with%20my%20question%20about%20this%20Coralogix%20documentation%20page.)[Open in Claude](https://claude.ai/new?q=Read%20https%3A%2F%2Fcoralogix.com%2Fdocs%2Fuser-guides%2Fapm-v2%2Ffeatures%2Ftraces.md%20and%20help%20me%20with%20my%20question%20about%20this%20Coralogix%20documentation%20page.) # Investigate traces and spans The **Traces** tab is the bridge from an aggregate metric - a latency spike or an error burst - to the individual requests behind it. It lists the entity's most recent spans and traces, with **Highlights**, **Spans**, and **Traces** views, a set of filters, and a chart of span volume over time. It is available for both services and databases. Use it to: * **Find the request behind a spike**: jump from an aggregate latency or error metric to the individual spans and traces that produced it. * **Inspect one request end to end**: open a trace to follow it across services and see where the time or the error occurred. * **Narrow to what matters**: filter by dataset, errored spans only, span kind, or a Lucene query to isolate the spans you care about. * **Open the full set in Explore**: send the current scope and filters to Explore when you need more than the most recent rows or the complete span body. [![The Traces tab: the filters bar (dataset, Errors toggle, span kind, and a Lucene query), a Count spans over time chart, and the Spans/Traces table.](/docs/assets/images/traces-tab-18bb1887144344a8811ce97504678aff.webp)](https://coralogix.com/docs/assets/images/traces-tab-18bb1887144344a8811ce97504678aff.webp) ## What you need[​](#what-you-need "Direct link to What you need") * Coralogix [Application Performance Monitoring (APM)](https://coralogix.com/docs/user-guides/apm-v2/getting-started/apm-onboarding-tutorial.md) installed and configured. * Spans reaching Coralogix for the entity you open. The tab reads spans directly, so it needs no additional processor or Span Metrics setup. ## Access the Traces tab[​](#access-the-traces-tab "Direct link to Access the Traces tab") 1. In your Coralogix toolbar, select **APM**. 2. Select a service or database to open its drilldown, then select the **Traces** tab. 3. The tab opens on the **Spans** view. Use the filters to scope what the chart and the views show, and the view toggle to switch between **Highlights**, **Spans**, and **Traces**. When you collapse the Traces section, its header keeps a summary of two mini-metrics - **Spans** and **Errors** - with the totals for the selected range. ## Filters[​](#filters "Direct link to Filters") A filters card at the top of the tab scopes the chart and all views together: * **Dataset**: the spans dataset to read. Defaults to the default spans dataset; select another to query a different one. * **Errors**: a toggle that restricts every view to errored spans. * **Span kind**: a multi-select, labelled **All kinds** when nothing is chosen. Filter to one or more of **Server**, **Client**, **Internal**, **Producer**, or **Consumer**. Selecting every kind is the same as selecting none. * **Lucene query** and **Run**: write a query in the search box ("Write a Lucene query to search your data...") and select **Run** to apply it. The query is applied on **Run**, not as you type. * **Explore Spans**: opens the current scope and filters in [Explore](https://coralogix.com/docs/user-guides/data_exploration/spans.md) in a new tab, without the row cap: use it for the full span body or for more than the most recent rows. ## Count spans over time[​](#count-spans-over-time "Direct link to Count spans over time") A chart titled **Count spans over time** plots the entity's span volume across the selected range as a single series. Use its menu to change the chart type and scale, set the time bucket, or open **View query**. Over a wide range the chart can exceed the query scan limit, in which case a warning appears: *"Counts are incomplete. This time range exceeded the query scan limit. Narrow the time range for exact numbers."* Narrow the range for exact counts. ## Spans and traces[​](#spans-and-traces "Direct link to Spans and traces") The tab lists the entity's recent activity in views toggled at the top: **Highlights**, **Spans** (the default), and **Traces**. A badge next to the toggle shows how many rows are loaded, and each table view paginates at 25 rows per page. Switching views refetches for the current time range and filters. ### Highlights[​](#highlights "Direct link to Highlights") The **Highlights** view breaks the entity's traces down by field and tag value, so you can see which values dominate its latency or errors without opening individual spans. Switch the breakdown between latency and errors, choose the aggregation, and turn on a compare-to-period mode to see what changed. A chart accompanies the breakdown; toggle it with **Show graph** / **Hide graph**. Note Each view shows the most recent rows only - up to **250 spans** or **100 traces**. When a view is capped, the row-count badge's tooltip reads, for example, *"Showing the most recent 250 spans. Open Explore for the full set."* Select **Explore Spans** to open the complete, uncapped set. ### Spans[​](#spans "Direct link to Spans") The **Spans** view lists individual spans, most recent first: * **#**: the row number. * **Timestamp**: when the span started. * **Operation**: the span's operation name. * **Span kind**: server, client, internal, producer, or consumer. * **Resource**: the instrumented resource that produced the span. * **Duration**: how long the span took. * **Status**: the span's outcome: success or error. ### Traces[​](#traces "Direct link to Traces") The **Traces** view rolls spans up into whole traces, most recent first: * **#**: the row number. * **Timestamp**: when the trace started. * **Operation**: the trace's entry-point operation. * **Resource**: the root resource the trace entered through. * **Span count**: total spans in the trace. * **Errors**: errored spans in the trace. * **Duration**: the trace's end-to-end duration. * **Breakdown**: how the trace's span time splits across services: this entity's own share versus everything downstream. ## Investigate a trace or span[​](#investigate-a-trace-or-span "Direct link to Investigate a trace or span") Select a row to open the **tracing drilldown**, which shows the full trace waterfall and each span's attributes. Selecting a span focuses that span within its trace; selecting a trace opens it from the root. See [Investigate with the tracing drilldown](https://coralogix.com/docs/user-guides/data_exploration/spans/span-drilldown.md) for the full reference. [![The tracing drilldown opened from a Traces row: the trace waterfall with each span\'s duration and share of the trace, and the span tree across services.](/docs/assets/images/trace-waterfall-509b24beb689a1c1c8111ec080c13c50.webp)](https://coralogix.com/docs/assets/images/trace-waterfall-509b24beb689a1c1c8111ec080c13c50.webp) ## Next steps[​](#next-steps "Direct link to Next steps") Profile a service's CPU and see where its time goes in [Profiling](https://coralogix.com/docs/user-guides/apm-v2/features/profiling.md).