# ExternalGroup reference

BEL provides
[external workload management for Linkerd](/buoyant-enterprise-linkerd/2.18/features/mesh-expansion/) in the
form of a Kubernetes controller and a proxy harness daemon. The controller runs
on the cluster and allows for automated `ExternalWorkload` lifecycle and
configuration management. This is achieved with the use of an `ExternalGroup`
resource which defines a template for creating `ExternalWorkload` resources.

### spec.template

| Field                 | Description                                                                                                                          |
| --------------------- | ------------------------------------------------------------------------------------------------------------------------------------ |
| `metadata.annotation` | Annotations that will be present on the created resource.                                                                            |
| `metadata.labels`     | Labels that will be present on the created resource.                                                                                 |
| `ports`               | A list of ports that the workload exposes.                                                                                           |
| `probes`              | A list of probes that the harness uses to determine and report the health of the workload (currently a max of 1 probe is supported). |

### spec.template.port

| Field            | Description                                                                 |
| ---------------- | --------------------------------------------------------------------------- |
| `name`           | Optional name for the port. Each named port can be referred to by services. |
| `port`           | The numeric port.                                                           |
| `protocol.ports` | Protocol exposed by the port. Must be UDP or TCP. Defaults to TCP.          |

### spec.template.probe

| Field                 | Description                                                                                                |
| --------------------- | ---------------------------------------------------------------------------------------------------------- |
| `failureThreshold`    | Amount of consecutive failures before probe reports a liveness failure (defaults to 3).                    |
| `successThreshold`    | Minimum consecutive successes for a probe to be considered successful after it has failed (defaults to 1). |
| `initialDelaySeconds` | Initial delay before probe will be triggered (defaults to 0).                                              |
| `timeoutSeconds`      | Number of seconds after which a probe times out (defaults to 1).                                           |
| `periodSeconds`       | How often to perform a probe (defaults to 10).                                                             |
| `httpGet.host`        | Hostname to connect to.                                                                                    |
| `httpGet.path`        | Path to access on the HTTP server.                                                                         |
| `httpGet.port`        | Port on the HTTP server to route to.                                                                       |
| `httpGet.scheme`      | Scheme to use for connecting to the host.                                                                  |
| `httpGet.httpHeaders` | Custom headers to add to the probe request.                                                                |
