Skip to content

[ISSUE] Configuration drift in databricks_cluster when is_single_node is set to true #4360

Open
@alexott

Description

@alexott

The is_single_node represents a simpler way of configuring single node clusters, but it automatically adds necessary spark_conf and custom_tags values, leading to the configuration drift, like this:

  # databricks_cluster.single_node_new_1228_101700_uxdk2a63 will be updated in-place
  ~ resource "databricks_cluster" "single_node_new_1228_101700_uxdk2a63" {
      ~ custom_tags                  = {
          - "ResourceClass" = "SingleNode" -> null
        }
        id                           = "1228-102112-aco3jrk3"
      ~ spark_conf                   = {
          - "spark.databricks.cluster.profile" = "singleNode" -> null
          - "spark.master"                     = "local[*, 4]" -> null
        }
        # (19 unchanged attributes hidden)

        # (1 unchanged block hidden)
    }

for the following configuration:

resource "databricks_cluster" "single_node_new_1228_101700_uxdk2a63" {
  spark_version      = "15.4.x-scala2.12"
  single_user_name   = "[email protected]"
  runtime_engine     = "STANDARD"
  node_type_id       = "Standard_D4ds_v5"
  kind               = "CLASSIC_PREVIEW"
  is_single_node     = true
  data_security_mode = "SINGLE_USER"
  cluster_name       = "Single Node new (TF)"
  autotermination_minutes = 10
}

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions