grafana_datasource/v0

Usage

This relation interface describes the expected behavior of any charm claiming to be able to provide a grafana datasource.

In most cases, this will be accomplished using the grafana_source library, although charm developers are free to provide alternative libraries as long as they fulfill the behavioral and schematic requirements described in this document.

Direction

The grafana_datasource interface implements a provider/requirer pattern. The provider is a charm that implements a grafana datasource-compatible endpoint, and the requirer is a charm that is able to use such an endpoint to query the data.

The requirer is furthermore expected to share back to the provider:

  • a unique identifier assigned to the source. This can be used by the provider to share with other charms for data correlation and cross-referencing purposes.

  • a unique identifier for the grafana application itself.

        flowchart TD
    Provider -- DatasourceEndpoint --> Requirer
    Requirer -- [DatasourceUID,GrafanaUID] --> Provider

Behavior

The requirer and the provider need to adhere to a certain set of criteria to be considered compatible with the interface.

Provider

  • Is expected to expose a server implementing the grafana source HTTP API.

  • Is expected to communicate said endpoint URL over unit data, as each unit will expose its own server.

Requirer

  • Is expected to share back via application data a mapping from provider unit names to unique datasource IDs.

  • Is expected to share back via application data a unique ID for the grafana application.

Relation Data

[Pydantic model]

Requirer

Additionally to a subset of the the (mandatory) juju topology fields, the requirer is expected to share the following fields:

  • type: the grafana datasource type. For the possible values see the upstream docs. Required.

  • extra_fields: used to configure certain datasources. Maps to the jsonData field. Optional.

  • secure_extra_fields: used to configure certain datasources. Maps to the secureJsonData field. Optional.

The whole configuration is expected to be json-encoded and nested under a grafana_source_data toplevel field.

Example

application_data: {
  grafana_source_data:
      {
        model: cos,
        model_uuid: 0000-0000-0000-0000,
        application: tempo,
        type: tempo,
        extra_fields: {
          some: value 
        },
        secure_extra_fields: {
          some: password 
        },

      } 
}

Provider

The provider is expected to share back a unique identifier for each unit of the requirer, as a mapping. This will be encoded as a json dict and nested under the datasource_uids field in the application databag. Also, is expected to share back a unique ID for the grafana application.

Example

application-data: {
 grafana_uid: 0000-0000-0000-0000,
 datasource_uids: {
  "tempo/0": 0000-0000-0000-0001, 
  "tempo/1": 0000-0000-0000-0002,
 }
}