[ISSUE] Issue with `databricks_cluster` resource

[manojhegde97]

workload_type configuration in databricks_cluster resource is not taking effect, we can see a blank in the cluster JSON . however when we use the true statement instead of false it is accepting,

Configuration

resource "databricks_cluster" "this" {
  provider                = databricks.ws
  cluster_name            = "manoj-job-new"
  spark_version           = data.databricks_spark_version.latest_lts.id
  node_type_id            = "c5d.xlarge"
  driver_node_type_id     = "c5d.xlarge"


  autoscale {
    min_workers = 1
    max_workers = 1
  }

  aws_attributes {
    zone_id = "auto"
  }

  # Prevent submitting job to run in interactive cluster
  workload_type {
    clients {
      jobs      = false
      notebooks = true
    }
  }

}

Expected Behavior

It should take the false statement for workload_type

Actual Behavior

only true statement is accepting

Terraform and provider versions

Terraform v1.6.2 , databricks : 1.47.0

[seanjw13]

I encountered this issue today.

With Databricks Terraform Provider v1.45.0 this correctly creates a cluster with workload tags set that prevent Jobs from running on it.

But when I update the Databricks provider to v1.46.0 or later, jobs=false is not propagated to the created cluster.

From looking at the release notes for v1.46.0, it seems like it may have to do with this change.

[Woffendm]

This is causing issues at my company as well.

[cpanpalia]

this is not fixed, I just tried with 1.52.0

resource "databricks_cluster" "deng_uc_cluster" { provider = databricks.created_workspace cluster_name = "DEng-UC-Cluster" spark_version = "14.3.x-scala2.12" node_type_id = "m6i.xlarge" driver_node_type_id = "m6i.xlarge" is_pinned = true data_security_mode = "USER_ISOLATION" runtime_engine = "STANDARD" num_workers = 0 enable_elastic_disk = false policy_id = <some_id>

enable_local_disk_encryption = false autotermination_minutes = 25

aws_attributes { first_on_demand = 1 ebs_volume_type = "GENERAL_PURPOSE_SSD" ebs_volume_count = 1 ebs_volume_size = 128 zone_id = "auto" }

autoscale { min_workers = 1 max_workers = 10 }

workload_type { clients { jobs = false notebooks = true } } }

and the postman API response mentions