Create & Manage Variables

Overview

A variable is like a placeholder for a value used in queries. Once set up in Custom Dashboards, your dashboard panel’s queries will update to show data for that value.

Variables make dashboards more interactive and dynamic. Instead of putting specific names for servers, applications, or sensors in your queries, you can use variables instead. Once set up, you can easily switch between different data views. This means that if you have very complex, high-cardinality metrics, you can add variables to zoom in on specific values you care about, across widgets for all data types and regardless of the variable source type.

Variable types

Choose from any of the following variable types:

Getting started

Navigate to Dashboards > Custom Dashboards.

In a new or existing dashboard, click the Manage Variable icon { } in the upper left-hand corner of your dashboard. Then click + Variable.

General options

Define the following fields for all variable types:

Variable type

Select a variable type.

Variable name

Text entered in this field determines how the variable name will appear in your query, e.g, {{podvalues}}. No capitals or spaces are allowed.

Display name

[Optional] Text entered in this field will appear in the top bar selection as part of the drop-down, e.g., “My Pods”.

Selection options

Once you have defined your variable, you may choose from these selection options:

• Multi-value: Allows multiple selection of values within a specific variable. Note that some queries do not support multiple values, such as {{variablename}}. Choose this option only if your query supports multiple values.
• Include all: You may enable this option when the multi-value option is enabled. The query will select all available and future values for a specific timeframe, regardless of your specific selection.

When both options are selected, all variable options are selected by default, and one or more may be deselected. Opting out of both options will allow you to select one variable value at a time.

Create a static-value variable

STEP 1. Select static-values as variable type.

STEP 2. Define the general options.

STEP 3. Enter the variable values manually, separated by a comma. For example: 2024, 2025, 2026, 2027.

STEP 4. Click SAVE.

Create a logs-based query variable

STEP 1. Select query as variable type.

STEP 2. Define the general options.

STEP 3. For query options:

• Select logs as your source.
• Select values as your variable type.
• Define the field by selecting a log key you wish to query. A list of current variable values presented in alpha-numerical order for the field will appear in the Values Preview.
• Sort values. Select how to sort your data in the Values Preview.

STEP 4. Click SAVE.

Create a spans-based query variable

STEP 1. Select query as variable type.

STEP 2. Define the general options.

STEP 3. For query options:

• Select spans as your source.
• Select values as your variable type.
• Define the field by selecting a log key you wish to query. A list of current variable values presented in alpha-numerical order for the field will appear in the Values Preview.
• Sort values. Select how to sort your data in the Values Preview.

STEP 4. Click SAVE.

Create a metrics-based query variable

Metric names

STEP 1. Select query as variable type.

STEP 2. Select general options.

STEP 3. For query options:

• Select metrics as your source.
• Select metric names as your variable type.
• Metrics regex. Enter a regex expression to fetch the list of metrics you would like displayed. E.g., ^kube(.*) will produce all metric names that begin with kube.
• Variable value. Determines how the value will appear in your query. A list of current variable values for the field will appear in the Values Preview.
• Variable value display name. Define how the values will be displayed in the drop-down selection in the top bar.
• Sort values. Select how to sort your data in the Values Preview.

STEP 5. Click SAVE.

Metric labels

STEP 1. Select query as variable type.

STEP 2. Select variable and display name.

STEP 3. For query options:

• Select metrics as your source.
• Select metric labels as your variable type. Choose the label for which you want the values displayed.
• Metrics regex. Enter a regex expression to fetch the list of metric labels you would like displayed. E.g., ^kube(.*) will produce all metric names that begin with kube.
• Variable value. Determines how the value will appear in your query. A list of current variable values for the field will appear in the Values Preview.
• Variable value display name. Define how the values will be displayed in the drop-down selection in the top bar.
• Sort values. Select how to sort your data in the Values Preview.

STEP 5. Click SAVE.

Metric label values

STEP 1. Select query as variable type.

STEP 2. Define the general options.

STEP 3. For query options:

• Select metrics as your source.
• Select label values as your variable type.
• Metric name. Choose the metric from which you want to pull labels.
• Metric label. Choose the label for which you want the values displayed. Make the label dynamic by inserting a variable, e.g., Pod Name = {{pod}}.
• Label filters. Create one or more nested variables and an additional filter layer for your query. (e.g., If you’ve chosen all of cluster the labels for a specific metric, you may choose to filter the labels by cluster. Select a value that is {{podvalue}} or the hardcoded name of your cluster.
• Variable value. Determines how the value will appear in your query. A list of current variable values for the field will appear in the Values Preview.
• Variable value display name. Define how the values will be displayed in the drop-down selection in the top bar.
• Sort values. Select how to sort your data in the Values Preview.

STEP 5. Click SAVE.

Manage variables

Easily view and manage the variables from the Custom Dashboards upper toolbar.

Edit / Delete

To edit or delete a variable, select the variable display name to be rerouted to the variable setup. To delete, click on the trash icon.

Query with variables

Query with Lucene, PromQL, or DataPrime.

PromQL

Template variables

Use {{ variable_name }} in the widget’s query language.

The PromQL query sum(kube_pod_owner{pod=~"{{pod}}"}) is used to aggregate metrics related to Kubernetes pods. {pod=~"{{pod}}"} is a label selector that filters the kube_pod_owner metric for the specific pod whose name matches the variable’s value. The {{pod}} placeholder is a template variable, which replaces the actual pod name when the query is executed.

Predefined variables

You may use predefined variables in your PromQL query.

For example, you could use the __range variable in sum_over_time:

sum_over_time(kube_pod_owner[${__range}])

Lucene

For example, use product_id:{{pid}} where pid is the actual variable name (not the display name).

DataPrime

For example,

• source logs | filter {{severity}}.arrayContains($m.severity)
• source logs | filter arrayContains({{severity}}, $m.severity)

{{severity}} is a logs-based variable created using the severity metadata field.

Additional resources

Check out this instructional video to jump-start your variable setup.

Support

Need help?

Our world-class customer success team is available 24/7 to walk you through your setup and answer any questions that may come up.

Feel free to contact us via our in-app chat or by emailing [email protected].