error 422 INVALID_REQUEST when using data_source nutanix_user_group

[mccasian]

Nutanix Cluster Information

Terraform Version

2024-12-07T18:57:55.329+0100 [INFO] Terraform version: 1.5.7
2024-12-07T18:57:55.330+0100 [DEBUG] using github.com/hashicorp/go-tfe v1.26.0
2024-12-07T18:57:55.330+0100 [DEBUG] using github.com/hashicorp/hcl/v2 v2.16.2
2024-12-07T18:57:55.330+0100 [DEBUG] using github.com/hashicorp/terraform-svchost v0.1.0
2024-12-07T18:57:55.330+0100 [DEBUG] using github.com/zclconf/go-cty v1.12.2
2024-12-07T18:57:55.330+0100 [INFO] Go runtime version: go1.23.1

Affected Resource(s)

nutanix_user_group

Terraform Configuration Files

locals {
  # read authorization_policies from the file
  cluster_admin_yaml = yamldecode(file("${path.module}/../../../../../../data/iam_config/cluster_admin.yaml"))
  
  # Convert the list to a set of strings
  cluster_admin_user_groups = toset([
    for user_group in local.cluster_admin_yaml.iam_entities.segment[var.segment].user_groups : user_group.name
  ])
}

data "nutanix_user_group" "user_groups_cluster_admin" {
  for_each = local.cluster_admin_user_groups
  user_group_distinguished_name = each.value # the name as written in the yaml file
}

Debug Output

---[ REQUEST ]---------------------------------------
POST /api/nutanix/v3/user_groups/list HTTP/1.1
Host: <censored_url>:9440
User-Agent: nutanix/v3
Content-Length: 47
Accept: application/json
Authorization: Basic cnVoMTo0RVgmeTQ0b0tZMi0hLTg=
Content-Type: application/json
Accept-Encoding: gzip

{
"filter": "",
"kind": "user_group",
"length": 100
}
​......
later on in the logs, because the user is not found, Terraform is trying the following:
---[ REQUEST ]---------------------------------------
POST /api/nutanix/v3/user_groups/list HTTP/1.1
Host: <censored_url>:9440
User-Agent: nutanix/v3
Content-Length: 52
Accept: application/json
Authorization: Basic cnVodo0RVgmeTQ0gMi0r33TTa=
Content-Type: application/json
Accept-Encoding: gzip

{
"filter": "",
"kind": "user",
"length": 100,
"offset": 0
}

Of course that fails, and the output is:
2024-12-07T18:25:59.051+0100 [ERROR] vertex "module.authorization_policies.data.nutanix_user_group.user_groups_vm_admin["cn=suser0021-l,ou=BetriebsGroups,ou=comp,ou=adGroups,dc=ad,dc=domainname,dc=com"]" error: error reading group with error error: {
"api_version": "3.1",
"code": 422,
"message_list": [
{
"details": {
"kind": [
"'user' is not one of ['user_group']"
]
},
"message": "Request could not be processed.",
"reason": "INVALID_REQUEST"
}
],
"state": "ERROR"
}

References

#697

I am not sure why exactly this happens, but since this is only happening on a single PC where we have a ton of users, I assume that the initial API call being limited to 100 results by default, is not able to find the user_group which is looking for. I've manually tested with a higher enough length, and the user_group can be found, but not with length 100.

I guess because the user_group cannot be found, terraform will try that with kind: user instead user_group ... but that can't work.

Best regards
Casian