Metrics
Overview
The APM > Metrics section displays metric data points collected from your applications and services.
A metric is a measurement (e.g., request duration, number of processed orders, or memory usage). Applications send their values as data points that contain the value (or aggregated values) measured at a specific time. A metric is similar to a Zabbix item, and a data point is similar to a value of that item.
A data point can be one of the following types:
- Gauge - a value measured at a specific time (e.g., current memory usage).
- Sum - a sum of measurements over a time period (e.g., total number of requests).
- Histogram - the number of measurements over a time period that fall into each of several fixed value ranges, called buckets (e.g., how many requests took 0-100 ms, 100-250 ms, or 250-500 ms).
- Exponential histogram - like a histogram, but each bucket is wider than the previous one by the same factor (e.g., 1-2 ms, 2-4 ms, 4-8 ms).
Each data point is produced by a service (e.g., zabbix-frontend; named by the OpenTelemetry service.name resource attribute), and by an instrumentation scope, the part of the code that created it (e.g., a library that measures HTTP request durations).
To provide context about each measurement, data points also have key-value attributes: data point attributes describe what was measured and distinguish data points of the same metric (e.g., http.request.method: POST), and resource attributes describe the service and the environment it runs in (e.g., host.name: eabf8c2a5ddc).
For more details on metrics, see the OpenTelemetry documentation.
The APM > Metrics section contains the following elements:
- Kiosk mode button - display only the page content, without the menu and header.
- Time period selector - select the time period for which to display data points.
- Filter - display only the data points you are interested in.
- Metric list - view the data points.
- Metric details - view the details of the selected data point.
Metric list
Each row in the metric list represents a data point.
Data points are displayed in a customizable table.
The 
Customize table button in the table header lets you show, hide, or reorder columns, and reset the layout to default.
You can also resize each column by dragging its right border; double-click the border to reset.
All customizations are applied immediately and remembered for the current user.
The metric list can have the following columns:
Each column, except Type, corresponds to a column in the metric tables (otel_metrics_gauge, otel_metrics_sum, otel_metrics_histogram, and otel_metrics_exponential_histogram) of the telemetry data source; for the list of columns, see Supported telemetry data columns.
Most columns have the same name as in the data source (e.g., Metric name corresponds to the MetricName column).
Where the names differ, the data source column is specified in the table below.
| Column | Description |
|---|---|
| Metric name | Metric that the data point belongs to (e.g., http.server.duration). |
| Type | Type of the data point: Gauge, Sum, Histogram, or Exponential histogram. Determined by the metric table that holds the data point. |
| Unit | Unit, if any, of the metric values (e.g., s or By; see UCUM).Data source column: MetricUnit. |
| Service name | Service that produced the data point. |
| Scope name | Instrumentation scope that produced the data point. |
| Start time | Date and time when the measurement period of the data point started. Data source column: StartTimeUnix. |
| Sum/Value | Sum of all measurements (Histogram, Exponential histogram), or the value (Gauge, Sum) of the data point. |
| Count | Number of measurements in a Histogram or Exponential histogram data point. |
| Time | Date and time of the data point; for Sum, Histogram, and Exponential histogram data points, the end of the measurement period. Data source column: TimeUnix. |
| Flags | Flags of the data point: 0 - no flags are set. 1 - no value was recorded for the data point. |
| Aggregation temporality | How the values of a Sum, Histogram, or Exponential histogram data point are aggregated over time: Unspecified - not specified. Delta - aggregated over the period since the previous data point. Cumulative - aggregated over the period since the start time. |
| Metric description | Description, if any, of the metric. |
| Attributes | Data point attributes. Click the  three-dot button to display all attributes. |
| Resource attributes | Resource attributes of the data point. Click the three-dot button to display all attributes. |
| Scope attributes | Attributes of the instrumentation scope. Click the three-dot button to display all attributes. |
Metric details
Click a data point in the metric list to open the metric details panel, which shows all information about the data point.
The metric details panel can have the following sections:
| Section | Description |
|---|---|
| Basic information | Metric name, type, unit, service name, and start time. |
| Histogram | Buckets (e.g., > 5 - 10), each with the number of measurements in it.Displayed only for Histogram data points. |
| Exponential histogram | Buckets (e.g., >= 249.7885 - 249.7887), each with the number of measurements in it.Displayed only for Exponential histogram data points. |
| Metric information | Values of the data point, which depend on its type, and the metric description, time, flags, and resource schema URL. |
| Attributes | Data point attributes. |
| Resource attributes | Resource attributes. |
| Scope attributes | Attributes, if any, of the instrumentation scope. |
| Scope | Name, version, and schema URL of the instrumentation scope. |
Using filter
You can use the filter to display only the data points you are interested in.
To filter by attributes, enter the attribute key, select the operator, and enter the attribute value.
The following operators are supported:
- Contains - (default) the attribute value contains the entered string.
- Equals - the attribute value equals the entered string.
- Does not contain - the attribute value does not contain the entered string.
- Does not equal - the attribute value does not equal the entered string.
- Exists - the attribute with the entered key exists.
- Does not exist - the attribute with the entered key does not exist.
If you add several conditions, select how to evaluate them:
- And/Or - (default) all conditions must be met; conditions with the same attribute key are combined with Or.
- Or - at least one condition must be met.