Made traffic_split field use default_from_api (#12323) (#20350)

[upstream:ab80e5eff2c6c9f9515622b93cdade13a4719e5c]

Signed-off-by: Modular Magician <[email protected]>

Author: modular-magician

.changelog/12323.txt

@@ -0,0 +1,3 @@
+```release-note:bug
+vertexai: fixed issue with google_vertex_ai_endpoint where upgrading to 6.11.0 would delete all traffic splits that were set outside Terraform (which was previously a required step for all meaningful use of this resource).
+```

google/services/vertexai/resource_vertex_ai_endpoint.go

@@ -192,17 +192,17 @@ Please refer to the field 'effective_labels' for all of the labels present on th
 			},
 			"traffic_split": {
 				Type:         schema.TypeString,
+				Computed:     true,
 				Optional:     true,
 				ValidateFunc: validation.StringIsJSON,
 				StateFunc:    func(v interface{}) string { s, _ := structure.NormalizeJsonString(v); return s },
 				Description: `A map from a DeployedModel's id to the percentage of this Endpoint's traffic that should be forwarded to that DeployedModel.
 If a DeployedModel's id is not listed in this map, then it receives no traffic.
-The traffic percentage values must add up to 100, or map must be empty if the Endpoint is to not accept any traffic at a moment.
-
-~> **Note:** The 'traffic_split' setting only applies after a model has been deployed to the endpoint. Re-applying a 'google_vertex_ai_endpoint'
-resource without updating the 'traffic_split' post-deployment may lead to your deployed 'traffic_split' being lost; see
+The traffic percentage values must add up to 100, or map must be empty if the Endpoint is to not accept any traffic at a moment. See
 the 'deployModel' [example](https://cloud.google.com/vertex-ai/docs/general/deployment#deploy_a_model_to_an_endpoint) and
-[documentation](https://cloud.google.com/vertex-ai/docs/reference/rest/v1beta1/projects.locations.endpoints/deployModel) for details.`,
+[documentation](https://cloud.google.com/vertex-ai/docs/reference/rest/v1beta1/projects.locations.endpoints/deployModel) for more information.
+
+~> **Note:** To set the map to empty, set '"{}"', apply, and then remove the field from your config.`,
 			},
 			"create_time": {
 				Type:        schema.TypeString,

website/docs/r/vertex_ai_endpoint.html.markdown

@@ -177,11 +177,10 @@ The following arguments are supported:
   (Optional)
   A map from a DeployedModel's id to the percentage of this Endpoint's traffic that should be forwarded to that DeployedModel.
   If a DeployedModel's id is not listed in this map, then it receives no traffic.
-  The traffic percentage values must add up to 100, or map must be empty if the Endpoint is to not accept any traffic at a moment.
-  ~> **Note:** The `traffic_split` setting only applies after a model has been deployed to the endpoint. Re-applying a `google_vertex_ai_endpoint`
-  resource without updating the `traffic_split` post-deployment may lead to your deployed `traffic_split` being lost; see
+  The traffic percentage values must add up to 100, or map must be empty if the Endpoint is to not accept any traffic at a moment. See
   the `deployModel` [example](https://cloud.google.com/vertex-ai/docs/general/deployment#deploy_a_model_to_an_endpoint) and
-  [documentation](https://cloud.google.com/vertex-ai/docs/reference/rest/v1beta1/projects.locations.endpoints/deployModel) for details.
+  [documentation](https://cloud.google.com/vertex-ai/docs/reference/rest/v1beta1/projects.locations.endpoints/deployModel) for more information.
+  ~> **Note:** To set the map to empty, set `"{}"`, apply, and then remove the field from your config.
 
 * `labels` -
   (Optional)