GraphQL Data Source

Grafana datasource plugin that provides access to a GraphQL API for numerical timeseries data, general/tabular data, annotations, and dashboard variables.

• The GraphQL query must be structured so that the data of interest is returned under the configurable data path (default data) in the response. If the object at that path is an array it will be iterated over, with each object added as a row in the data frame, otherwise the result object will be the only row.
  • Can be separated by commas to use multiple data paths
• Timeseries data must include a timestamp field under the data path, default Time, in ISO8601 or a configurable custom format.
• Nested types will be flattened into dot-delimited fields.
• Grafana variables should be substituted directly in the query (instead of using GraphQL variables). The dashboard time ranges are available in the global variables $__from and $__to as millisecond epoch (or in whatever format is needed by the API in Grafana 7.1.2 or later).
• Group by can be used to group elements into multiple data points.
• Alias by is used to alter the name of the field displayed in the legend. $field_<field.name> is substituted with the values of the field and $fieldName is substituted with the name of the field.

Screenshots

Examples

Below are some example queries demonstrating how to use the plugin, using the FIFEMon GraphQL test source server, which also includes a dashboard demonstrating these queries.

Basic timeseries

query {
  data: simple_series(from: "${__from:date:iso}", to: "${__to:date:iso}", interval_ms: $__interval_ms) {
    Time: timestamp
    value
  }
}

Note the use of the global $__from and $__to variables to insert the dashboard time range into the query and the use of $__interval_ms to specify the appropriate time interval for the graph.

Custom time format

query {
  simple_series(
    from: "${__from:date:iso}"
    to: "${__to:date:iso}"
    interval_ms: $__interval_ms
    format: "MM.dd.uuuu HHmmss"
  ) {
    timestamp
    value
  }
}

• Data path: simple_series
• Time path: timestamp
• Time format: MM.DD.YYYY HHmmss

Alias and group by

query {
  complex_series(from: "${__from:date:iso}", to: "${__to:date:iso}", interval_ms: $__interval_ms) {
    time {
      timestamp
    }
    value
    group {
      id
      name
    }
  }
}

• Data path: complex_series
• Time path: time.timestamp
• Group by: group.id
• Alias by: $field_group.name

In the above example, "Group by" and "Alias by" are defined. "Group by" allows you to split up an array of data into multiple data points. "Alias by" is used as the name of the data point. You can make alias use text from the query or even the field name by using $field_<your.field.name> for the value of the field, or $fieldName for the name of the field. If $fieldName was used, it would be replaced by "value" because that's the name of the field. If $field_group.name was used, it would be replaced with the value of name. Using $fieldName can be useful if you're querying multiple numeric fields that you want displayed in your graph.

Annotations

query {
  events(from: "${__from:date:iso}", to: "${__to:date:iso}", end: true) {
    timestamp
    end_timestamp
    name
    description
    tags
  }
}

• Data path: events
• Time path: timestamp
• End time path: end_timestamp
• Title: $field_name
• Text: $field_description
• Tags: tag1, tag2

The above annotation example is similar to regular queries. You are able to define a data path, time path, and time format. Similar to the last example, you can also substitute values into the title, text, and tags by using $field_<field name>. Tags are separated by commas. The above example has two tags: "tag1" and "tag2".

If the optional end time field is defined and present, the annotation will be shown over a period of time.

Dashboard Variable Queries

Dashboard variables can be populated by a GraphQL query that returns an array of objects. If the objects contain both __text and __value fields then they will be used (the __text field will be displayed, the __value field will be used in substitutions). Otherwise the values of all fields will be appended to the variable value list.

query {
  groups {
    __value: id
    __text: name
  }
}

• Data path: groups