Query type

For the query type, select either an instant or range query. The query is run over a specific time frame that is defined by the dashboard time picker or the widget time frame.

• Instant query mode displays the metric's value at a single point in time, specifically at the end timestamp of the selected time frame. Instant query mode is ideal for real-time monitoring, providing a snapshot of data at the moment of visualization. It is also more performant, retrieving only the most recent data point without fetching the entire dataset.

• Range query mode retrieves and aggregates metric data over a specified time frame. It returns all data points between the start and end timestamps that match your query. As the time frame changes, the system automatically adjusts the interval to optimize the number of data points displayed in the graph. Range query mode is best used for identifying trends or evaluating performance over a defined period.

Metric name

Choose your preferred metric from the drop-down menu that is dynamically filled with available metrics. You can use the autocomplete capability to explore the metric name, labels, and values.

Filters

To add a filter, click the +. Choose a metric label and an associated value to filter your metric.

Select the = or != operator from the drop-down menu to include or exclude one or more values. By using the =~ or !~ operators, you can input a RegEx expression.

To add more label-value pairs, click the + to add a line.

Functions

By selecting metrics and labels create a valid Prometheus query, but you can also create more complex queries by using functions.

To add a function, click +. Select between aggregation, count, and rollup functions.

Aggregation functions calculate a set of values and return a single value. These functions include SUM(), AVG(), MIN(), MAX(), COUNT(), and others. After you choose the function, select the label to aggregate in the drop-down menu or enter a label value.

Count functions run calculations on a set of values and return a single value. These functions include Count(), Absent(), Absent over time(), Present over time(), Changes(), Resets(), and others.

A rollup function refers to functions that aggregate time series data over a specified time range. These functions compute metrics such as average (avg_over_time), sum (sum_over_time), minimum (min_over_time), maximum (max_over_time), count (count_over_time), and quantiles (quantile_over_time) over a window of time. They are used to summarize and analyze metric data.

After you choose the function, select the range to be queried as a number or ${__range}. ${__range} represents the duration of the dashboard time range. It is rendered as an interval string supported by PromQL. For example, if you select a time range from 13.00 to 14.30, the ${__range} variable is rendered as 90m.

With rank functions you can sort, rank, and filter data within your queries. You can use them to refine your results based on specific metrics and values.

To add a rank function, click + and choose from TOPK, SORT, or SORT Descending. After you select a rank function, configure it by specifying the metric and parameters (for example, populate the number of K results you would like the query to retrieve for TOPK).

Filters

To add a filter, click +. Choose a label and an associated value.

Select the = or != operator from the drop-down menu to include or exclude one or more values. By using the =~ or !~ operators, you can input a RegEx expression.

To add more label-value pairs, click the +.

Functions

To add a function, click +. Select an aggregated value by using a function.

With Group by, you can group query results by one or more fields.