[ISSUE] Issue with workspace provider on newly created workspaces

[kevin-zwartscholten]

Description

From provider version 1.114.0 and onward, I get errors on terraform plan/apply when I create a new workspace and resources in that workspace in the same deployment. Below an example with databricks_workspace_conf, but this error seems to happen for all resources created with a workspace provider.

Configuration

provider "databricks" {
  alias = "mws"
  host  = "https://accounts.cloud.databricks.com"
}

resource "databricks_mws_workspaces" "workspace" {
  account_id                  = var.databricks_account_id
  workspace_name              = "my-workspace"
  aws_region                  = "<region>"
  compute_mode                = "SERVERLESS"
}


provider "databricks" {
  alias = "workspace"
  host  = databricks_mws_workspaces.workspace.workspace_url
}

resource "databricks_workspace_conf" "this" {
  provider = databricks.workspace

  custom_config = {
    "maxTokenLifetimeDays" = "30"
  }
}

Expected Behavior

terraform plan and terraform apply should succeed without error

Actual Behavior

│ Error: managing workspace-level resources requires a workspace_id, but none was found in the resource's provider_config block or the provider's workspace_id attribute
│ 
│   with databricks_workspace_conf.this,
│   on main.tf line 19, in resource "databricks_workspace_conf" "this":
│   19: resource "databricks_workspace_conf" "this" {
│ 
╵

It seems to be the issue that the workspace_id does not exist yet, because even if I configure workspace_id on the provider or add provider_config block to the databricks_workspace_conf I get the same issue. If I deploy in two steps, first the databricks_mws_workspaces and then the databricks_workspace_conf I do not run into this issue.

Steps to Reproduce

terraform plan / terraform apply

Terraform and provider versions

1.114.0 this started happening, issue is still present in 1.121.0

Is it a regression?

Yes, this worked until 1.113.0

Debug Output

Error: -14T11:23:48.006Z [ERROR] provider.terraform-provider-databricks_v1.114.0: Response contains error diagnostic: tf_resource_type=databricks_workspace_conf tf_rpc=PlanResourceChange @caller=/home/runner/go/pkg/mod/github.com/hashicorp/terraform-plugin-go@v0.30.0/tfprotov6/internal/diag/diagnostics.go:58 @module=sdk.proto diagnostic_detail="" diagnostic_severity=ERROR diagnostic_summary="managing workspace-level resources requires a workspace_id, but none was found in the resource's provider_config block or the provider's workspace_id attribute" tf_proto_version=6.11 tf_provider_addr=registry.terraform.io/databricks/databricks tf_req_id=513fc9cd-a95e-e399-d49f-7cff79caaf3c timestamp=2026-07-14T11:23:48.006Z
Error: -14T11:23:48.006Z [ERROR] vertex "databricks_workspace_conf.this" error: managing workspace-level resources requires a workspace_id, but none was found in the resource's provider_config block or the provider's workspace_id attribute
Error: -14T11:23:48.006Z [ERROR] vertex "databricks_workspace_conf.this (expand)" error: managing workspace-level resources requires a workspace_id, but none was found in the resource's provider_config block or the provider's workspace_id attribute

Important Factoids

Would you like to implement a fix?

[lmal54321]

This is not a bug - it's a core terraform principal

https://developer.hashicorp.com/terraform/language/block/provider

You can use [expressions](https://developer.hashicorp.com/terraform/language/expressions) to configure provider arguments, but you can only reference values that Terraform knows before it applies your configuration. You can reference input variables and arguments that you specify directly in your configuration, but you cannot reference computed resource attributes, such as google.web.public_ip.