Dash0 Acquires Lumigo to Expand Agentic Observability Across AWS and Serverless

Last updated: March 3, 2026

Add Variables

Overview

Variables are an advanced feature that enables dynamic filtering across dashboard panels.

A variable defines a selectable parameter—such as a service name attribute—that is substituted into PromQL queries at runtime.

Create a Variable

  1. Click Add variable in the dashboard toolbar.

    Create dashboard

    The Add new variable dialog opens.

    Variables

  2. Select the attribute to use as the variable source (e.g., service.name).

  3. Configure the variable's display name and default value.

Use Variables in Queries

Once defined, reference the variable in any panel's PromQL query using the standard variable substitution syntax. When a user changes the variable's selection, all panels referencing that variable update automatically.

Variables

Understand Variable Types

Dash0 has three different ways of using dashboard variables:

Text Variables

In Dash0, dashboarding variables can represent variables containing text that is defined once and can be reused in every panel.

That text can be used as a simple replacement in queries, for example:

promql
1
spans_total{my_label=~"$variable_name"}

List Variables

Furthermore, variables can be lists of values. These lists can be a static source, a list based on time-series label values, a list of time-series label names, or a result of a PromQL query.

In this case, a filter will be specified by a list instead of a single value:

promql
1
spans_total{my_label=~"$variable_name"}

Filter Variables

The most powerful capability of variables in Dash0 is their use as filters. A filter variable name can be specified with an attribute key. The specified attribute key is statically set, but the operator, as well as the values to match, can be flexibly defined by the user of a dashboard:

promql
1
spans_total{$variable_name}

Filter variables support hierarchical dependencies, enabling chained filtering. Dependent variables dynamically populate options based on parent selections, intelligently adapting operators. The system provides user feedback when constraints result in no available values. Configure dependencies via the "Depends on filter variables" setting to establish parent-child relationships for drill-down data exploration.