Add and manage variables

The following table lists the types of variables in Grafana.

Entering General options

You must enter general options for any type of variable that you create.

Adding a query variable

Query variables enable you to write a data source query that can return a list of metric names, tag values, or keys. For example, a query variable might return a list of server names, sensor IDs, or data centers. The variable values change as they dynamically fetch options with a data source query.

Query variables are generally only supported for strings. If your query returns numbers or any other data type, you might need to convert them to strings in order to use them as variables. For the Azure data source, for example, you can use the tostring function for this purpose.

Query expressions can contain references to other variables and in effect create linked variables. Grafana detects this and automatically refreshes a variable when one of its linked variables change.

Adding a custom variable

Use a custom variable for a value that does not change, such as a number or a string.

For example, if you have server names or Region names that never change, then you might want to create them as custom variables rather than query variables. Because they do not change, you might use them in chained variables rather than other query variables. That would reduce the number of queries Grafana must send when chained variables are updated.

Adding a text box variable

Text box variables display a free text input field with an optional default value. This is the most flexible variable, because you can enter any value. Use this type of variable if you have metrics with high cardinality or if you want to update multiple panels in a dashboard at the same time.

Adding a constant variable

Constant variables enable you to define a hidden constant. This is useful for metric path prefixes for dashboards you want to share. When you export a dashboard, constant variables are converted to import options.

Constant variables are not flexible. Each constant variable only holds one value, and it cannot be updated unless you update the variable settings.

Constant variables are useful when you have complex values that you need to include in queries but don’t want to retype in every query. For example, if you had a server path called i-0b6a61efe2ab843gg, then you could replace it with a variable called $path_gg.

Adding a data source variable

Data source variables enable you to quickly change the data source for an entire dashboard. They are useful if you have multiple instances of a data source, perhaps in different environments.

Adding an interval variable

Use an interval variable to represents time spans such as 1m,1h, or 1d. You can think of them as a dashboard-wide group by time command. Interval variables change how the data is grouped in the visualization. You can also use the Auto Option to return a set number of data points per time span.

You can use an interval variable as a parameter to group by time (for InfluxDB), date histogram interval (for Elasticsearch), or as a summarize function parameter (for Graphite).

Interval variable examples

The following example shows a template variable myinterval in a Graphite function:

summarize($myinterval, sum, false)

Adding ad hoc filters

Ad hoc filters enable you to add key-value filters that are automatically added to all metric queries that use the specified data source. Unlike other variables, you do not use ad hoc filters in queries. Instead, you use ad hoc filters to write filters for existing queries.

Create ad hoc filters

Ad hoc filters are one of the most complex and flexible variable options available. Instead of a regular list of variable options, this variable allows you to build a dashboard-wide ad hoc query. Filters you apply in this manner are applied to all panels on the dashboard.

Configure variable selection options

Selection Options are a feature you can use to manage variable option selections. All selection options are optional, and they are off by default.

Multi-value variables

Interpolating a variable with multiple values selected is tricky as it is not straight forward how to format the multiple values into a string that is valid in the given context where the variable is used. Grafana tries to solve this by allowing each data source plugin to inform the templating interpolation engine what format to use for multiple values.

Multi-value variables with a Graphite data source

Graphite uses glob expressions. A variable with multiple values is, in this case, be interpolated as {host1,host2,host3} if the current variable value iss host1, host2, and host3.

Multi-value variables with a Prometheus or InfluxDB datasource

InfluxDB and Prometheus use regex expressions, so the same variable is interpolated as (host1|host2|host3). Every value is also regex escaped. If it were not, a value with a regex control character would break the regex expression.

Multi-value variables with an Elastic data source

Elasticsearch uses lucene query syntax, so the same variable is formatted as ("host1" OR "host2" OR "host3"). In this case, every value is escaped so that the value only contains lucene control words and quotation marks.

Troubleshoot multi-value variables

Automatic escaping and formatting can cause problems and it can be tricky to grasp the logic behind it. Especially for InfluxDB and Prometheus where the use of regex syntax requires that the variable is used in regex operator context.

If you do not want Grafana to do this automatic regex escaping and formatting, then you must do one of the following:

Include All option

Grafana adds an All option to the variable dropdown list. If a user selects this option, then all variable options are selected.

Custom all value

This option is only visible if the Include All option is selected.

Enter regex, globs, or lucene syntax in the Custom all value field to define the value of the All option.

By default the All value includes all options in a combined expression. This can become very long and can have performance problems. Sometimes it can be better to specify a custom all value, like a wildcard regex.

To have custom regex, globs, or lucene syntax in the Custom all value option, it is never escaped so you will have to think about what is a valid value for your data source.

Global variables

Grafana has global built-in variables that can be used in expressions in the query editor. This topic lists them in alphabetical order and defines them. These variables are useful in queries, dashboard links, panel links, and data links.

$__dashboard

This variable is the name of the current dashboard.

$__from and $__to

Grafana has two built-in time range variables: $__from and $__to. They are currently always interpolated as epoch milliseconds by default, but you can control date formatting.

The syntax above also works with ${__to}.

$__interval

You can use the $__interval variable as a parameter to group by time (for InfluxDB, MySQL, Postgres, MSSQL), Date histogram interval (for Elasticsearch), or as a summarize function parameter (for Graphite).

Grafana automatically calculates an interval that can be used to group by time in queries. When there are more data points than can be shown on a graph, the queries can be made more efficient by grouping by a larger interval. For example, if you are looking at a graph of 3 months worth of data, you might not be able to see detail at the minute level. Grouping by the hour or day makes the query more efficient without affecting what the graph shows. The $__interval is calculated using the time range and the width of the graph (the number of pixels).

Approximate Calculation: (to - from) / resolution

For example, when the time range is 1 hour and the graph is full screen, then the interval might be calculated to 2m - points are grouped in 2 minute intervals. If the time range is 6 months and the graph is full screen, then the interval might be 1d (1 day) - points are grouped by day.

In the InfluxDB data source, the legacy variable $interval is the same variable. $__interval should be used instead.

The InfluxDB and Elasticsearch data sources have Group by time interval fields that are used to hard code the interval or to set the minimum limit for the $__interval variable (by using the > syntax, for example >10m).

$__interval_ms

This variable is the $__interval variable in milliseconds, not a time interval formatted string. For example, if the $__interval is 20m then the $__interval_ms is 1200000.

$__org

This variable is the ID of the current organization. ${__org.name} is the name of the current organization.

$__user

${__user.id} is the ID of the current user. ${__user.login} is the login handle of the current user. ${__user.email} is the email for the current user.

$__range

Only supported for Prometheus and Loki data sources. This variable represents the range for the current dashboard. It is calculated by to - from. It has a millisecond and a second representation called $__range_ms and $__range_s.

$__rate_interval

Only supported for Prometheus data sources. The $__rate_interval variable is meant to be used in the rate function.

$timeFilter or $__timeFilter

The $timeFilter variable returns the currently selected time range as an expression. For example, the time range interval Last 7 days expression is time > now() - 7d.

This is used in several places, including:

$__timezone

The $__timezone variable returns the currently selected time zone, either utc or an entry of the IANA time zone data base (for example, America/New_York).

If the currently selected time zone is Browser Time, Grafana will try to determine your browser time zone.

Chained variables

Chained variables, also called linked variables or nested variables, are query variables with one or more other variables in their variable query.

Chained variable queries are different for every data source, but the premise is the same for all. You can use chained variable queries in any data source that allows them.

Extremely complex linked templated dashboards are possible, 5 or 10 levels deep. Technically, there is no limit to how deep or complex you can go, but the more links you have, the greater the query load.

Best practices and tips

The following practices will make your dashboards and variables easier to use.

Creating new linked variables

Variable order

You can change the orders of variables in the dashboard variable list by clicking the up and down arrows on the right side of each entry. Grafana lists variable dropdowns left to right according to this list, with the variable at the top on the far left.

Complexity consideration

The more layers of dependency you have in variables, the longer it will take to update dashboards after you change variables.

For example, if you have a series of four linked variables (country, Region, server, metric) and you change a root variable value (country), then Grafana must run queries for all the dependent variables before it updates the visualizations in the dashboard.

Manage variables

The variables page lets you add variables and manage existing variables. It also allows you to inspect variables and identify whether a variable is being referenced (or used) in other variables or dashboard.

Move – You can move a variable up or down the list using drag and drop.

Clone – To clone a variable, click the clone icon from the set of icons on the right. This creates a copy of the variable with the name of the original variable prefixed with copy_of_.

Delete – To delete a variable, click the trash icon from the set of icons on the right.

Filter variables with regex

Using the Regex Query option, you filter the list of options returned by the variable query or modify the options returned.

This page shows how to use regex to filter/modify values in the variable dropdown.

Using the Regex Query Option, you filter the list of options returned by the Variable query or modify the options returned. For more information, refer to the Mozilla guide on Regular expressions.

The following examples show filtering on the following list of options

backend_01
backend_02
backend_03
backend_04

Filter so that only the options that end with 01 or 02 are returned

Regex:

/
(
01|02 
) 
$/

Result:

backend_01
backend_02

Filter and modify the options using a regex capture group to return part of the text

Regex:

/.* 
(
01|02 
)
/

Result:

01
02

Filter and modify - Prometheus Example

For this list of options:

up{instance="demo.robustperception.io:9090",job="prometheus"} 1 1521630638000
up{instance="demo.robustperception.io:9093",job="alertmanager"} 1 1521630638000
up{instance="demo.robustperception.io:9100",job="node"} 1 1521630638000

This regex:

/. *instance="
(
[^"]*
)
.*/

Returns these results:

demo.robustperception.io:9090
demo.robustperception.io:9093
demo.robustperception.io:9100

Filter and modify using named text and value capture groups

Using named capture groups, you can capture separate ’text’ and ‘value’ parts from the options returned by the variable query. This allows the variable dropdown list to contain a friendly name for each value that can be selected.

For example, when querying the node_hwmon_chip_names Prometheus metric, the chip_name is a lot friendlier that the chip value. So the following variable query result:

node_hwmon_chip_names{chip="0000:d7:00_0_0000:d8:00_0",chip_name="enp216s0f0np0"} 1
node_hwmon_chip_names{chip="0000:d7:00_0_0000:d8:00_1",chip_name="enp216s0f0np1"} 1
node_hwmon_chip_names{chip="0000:d7:00_0_0000:d8:00_2",chip_name="enp216s0f0np2"} 1
node_hwmon_chip_names{chip="0000:d7:00_0_0000:d8:00_3",chip_name="enp216s0f0np3"} 1

Passed through the following Regex:

/chip_name="(?<text>[ ^ " ] + ) |chip=" (?<value >[ ^ " ] + )/g

Would produce the following dropdown list:

Display Name          Value
------------          -------------------------
enp216s0f0np0         0000:d7:00_0_0000:d8:00_0
enp216s0f0np1         0000:d7:00_0_0000:d8:00_1
enp216s0f0np2         0000:d7:00_0_0000:d8:00_2
enp216s0f0np3         0000:d7:00_0_0000:d8:00_3

Only text and value capture group names are supported.