# Trace Details

To see the details of the trace, click the row that lists the trace.

The trace information is both interactive and detailed, and consists of these main areas:

- [Flame graph](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#flame)
- [Span List](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#span)
- [Waterfall](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#waterfall)

## Flame graph

The [Flame graph](https://docs.kloudfuse.com/platform/3.2.5/glossary/#flame-graph) breaks out the trace into the component spans and illustrates which services impact the performance, both because of completion time and because of errors.

The chart provides the following information and affordances:

1. The **Trace id**.
2. The **Total Duration** of the trace.
3. The **Latency** metric of the resource.

Hover your pointer over the latency icon; Kloudfuse displays the latency metric in relation to the distribution of the resource.

4. The **Start time** of the trace.
5. The **Waterfall Graph** itself, with components identified by name, start time, duration, error status, and type of service or span.
6. Hover your pointer over one of the components to see its **individual details**: _name_, _status_, _duration_, _percentage_ of total time, and _span kind/type_.
7. When you select different components of the graph, the reporting changes to show the information for that service.
8. By default, the right side of the interface shows all **Services** involved in the trace.

You change the display between **Execution time (%)** and **Span Count**.

You can also switch between viewing **Services** and **Hosts**. Here, too, you have both **Execution time (%)** and **Span Count** options.

9. The trace detail posts the information on the specifics of the selected service within the span: _total time_, _percentage_ of total execution time, and the _latency_ (P value). Hover over the latency graph icon to see the details.

Depending on the type of service and its status, trace detail also reports on the following [Supplemental details](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#supplemental): [Info](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#info), [Metrics](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#metrics), [Logs](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#logs), and [Stacktrace](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#stacktrace).

## Span List

The **Span List** interface groups all spans by [service](https://docs.kloudfuse.com/platform/3.2.5/glossary/#service) and [span type](https://docs.kloudfuse.com/platform/3.2.5/glossary/#span-type), and reports on their **Name**, number of **Spans**, **Duration**, **Exec Time**, and **Exec %**:

Click the services and span type groupings to expand them, and see individual spans:

Select spans to see their individual [Supplemental details](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#supplemental): [Info](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#info), [Metrics](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#metrics), [Logs](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#logs), and [Stacktrace](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#stacktrace).

If you plan to examine the spans by type, use the quick filters above the list:

for custom spans

for database spans

for web spans

for queue spans

for gRPC spans

To simplify finding spans in the list, use the **Search spans** option above the list.

## Waterfall

The **Waterfall** view of a trace combines many of the visualization features of the [Flame graph](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#flame) with the category reporting of the [Span List](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#span).

As you select the different services in the waterfall interface, Kloudfuse displays the relevant [Supplemental details](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#supplemental).

Note also that you have options for examining the **[Service](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#service)** and **[Host](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#host)** collections, like in the **[Flame graph](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#flame)** interface.

## Supplemental information

In addition to Flame graph, etc., Kloudfuse reports these supplemental details for the trace, or a span: [Info](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#info), [Metrics](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#metrics), [Logs](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#logs), and [Stacktrace](https://docs.kloudfuse.com/platform/3.2.5/apm-trace-detail/#stacktrace).

The following sections explain the supplemental data for a **frontend** service.

### Info

The **Info** is the JSON definition of the span, including its attributes, duration, endpoints, methods, names, parent span, service, and many more.

If you hover over any element of the Info report, a  (copy icon appears); use it to copy relevant information.

### Metrics

The **Metrics** section reports the measurements for the **Service** in the trace or span using the following graphs:

- **Hits** (Requests/sec or number of Requests);
- **Errors** (Errors/sec, Errors, or Errors %);
- **Latency** that you can select from the drop-down;
- **Execution time breakdown**.

If you select **Endpoint** instead of **Service**, you will see similar graphs. The **Execution time breakdown** chart does not appear for endpoints.

### Logs

The **Logs** section lists the logs that match the trace ID.

#### Match other attributes

You can choose to view the logs by matching availability zone, kube cluster name, kube namespace, or pod name:

#### Show optional fields

Use the  menu to select what fields to show in the log list, and how many log lines to show:

#### Download logs

Kloudfuse downloads _all_ log data, not just the portions you choose to display.

Click the  (Download) button, and select one of the format options by clicking on it.

The download starts immediately.

### Stacktrace

The stacktrace is a report of the active stack frames; it helps you debug errors in your system.