node_storage_quota_in_mb can't be set to the document suggested values for azurerm_cosmosdb_postgresql_cluster

leilifremont commented 12 months ago

Is there an existing issue for this?

[X] I have searched the existing issues

Community Note

Please vote on this issue by adding a :thumbsup: reaction to the original issue to help the community and maintainers prioritize this request
Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
If you are interested in working on this issue or have submitted a pull request, please leave a comment and review the contribution guide to help.

Terraform Version

1.5.5

AzureRM Provider Version

3.76.0

Affected Resource(s)/Data Source(s)

azurerm_cosmosdb_postgresql_cluster

Terraform Configuration Files

resource "azurerm_resource_group" "rg" {
  name     = "testrg"
  location = "eastus"
}

resource "random_password" "cosmosdb_postgresql_passwords" {
  length           = 24
  min_upper        = 6
  min_lower        = 4
  min_numeric      = 6
  special          = false

  depends_on = [
    resource.azurerm_resource_group.rg
  ]
}

resource "azurerm_cosmosdb_postgresql_cluster" "primary_replica" {
  name                            = "primary"
  location                        = "eastus"
  resource_group_name             = azurerm_resource_group.rg.name
  administrator_login_password    = random_password.cosmosdb_postgresql_passwords.result
  coordinator_storage_quota_in_mb = 262144
  coordinator_vcore_count         = 2
  node_count                      = 2
  node_storage_quota_in_mb        = 262144
  node_vcores                     = 2
  node_server_edition             = "MemoryOptimized"

  coordinator_public_ip_access_enabled = false
  citus_version                        = "12.1"
  sql_version                          = "16"

  depends_on = [
    random_password.cosmosdb_postgresql_passwords
  ]
}

resource "time_sleep" "wait_20_minutes" {
  create_duration = "1200s"
  triggers = {
    postgresql_cluster_id  = azurerm_cosmosdb_postgresql_cluster.primary_replica.id
  }
}

resource "azurerm_cosmosdb_postgresql_cluster" "read_replica" {
  name                            = "secondary"
  resource_group_name             = azurerm_resource_group.rg.name
  location                        = "westus3"
  source_location      = azurerm_cosmosdb_postgresql_cluster.primary_replica.location
  source_resource_id   = azurerm_cosmosdb_postgresql_cluster.primary_replica.id
  administrator_login_password    = random_password.cosmosdb_postgresql_passwords.result
  coordinator_storage_quota_in_mb = 262144
  coordinator_vcore_count         = 2
  node_count                      = 2
  node_storage_quota_in_mb        = 262144
  node_vcores                     = 2
  node_server_edition             = "MemoryOptimized"

  coordinator_public_ip_access_enabled = true
  citus_version                        = "12.1"
  sql_version                          = "16"

  depends_on = [
    time_sleep.wait_20_minutes
  ]
}

Debug Output/Panic Output

╷
│ Error: creating Server Groupsv 2 (Subscription: "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
│ Resource Group Name: "PeterTest10252023"
│ Server Groupsv 2 Name: "peter1025primary"): performing Create: unexpected status 400 with error: bad_request: Worker disk size of 262144 is not allowed.
│
│   with azurerm_cosmosdb_postgresql_cluster.primary_replica,
│   on main.tf line 47, in resource "azurerm_cosmosdb_postgresql_cluster" "primary_replica":
│   47: resource "azurerm_cosmosdb_postgresql_cluster" "primary_replica" {
│
│ creating Server Groupsv 2 (Subscription: "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
│ Resource Group Name: "PeterTest10252023"
│ Server Groupsv 2 Name: "peter1025primary"): performing Create: unexpected status 400 with error: bad_request: Worker disk size of 262144 is not allowed.
╵

Expected Behaviour

Expected resource can be created. Actually this parameter works in last week.

From document: node_storage_quota_in_mb - (Optional) The storage quota in MB on each worker node. Possible values are 32768, 65536, 131072, 262144, 524288, 1048576, 2097152, 4194304, 8388608 and 16777216.

Seems 32768, 65536, 131072, 262144 doesn't work in my case? Is this related to the Coordinator disk size?

Actual Behaviour

This failed ╷ │ Error: creating Server Groupsv 2 (Subscription: "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" │ Resource Group Name: "PeterTest10252023" │ Server Groupsv 2 Name: "peter1025primary"): performing Create: unexpected status 400 with error: bad_request: Worker disk size of 262144 is not allowed. │ │ with azurerm_cosmosdb_postgresql_cluster.primary_replica, │ on main.tf line 47, in resource "azurerm_cosmosdb_postgresql_cluster" "primary_replica": │ 47: resource "azurerm_cosmosdb_postgresql_cluster" "primary_replica" { │ │ creating Server Groupsv 2 (Subscription: "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" │ Resource Group Name: "PeterTest10252023" │ Server Groupsv 2 Name: "peter1025primary"): performing Create: unexpected status 400 with error: bad_request: Worker disk size of 262144 is not allowed. ╵

Steps to Reproduce

No response

Important Factoids

No response

References

No response

neil-yechenwei commented 12 months ago

Thanks for raising this issue. Seems it's supported by API. For more usage problem, suggest to file the issue on TF community form https://discuss.hashicorp.com/c/terraform-providers/tf-azure/34.

resource "azurerm_cosmosdb_postgresql_cluster" "test" {
  name                = "acctestclustertest02"
  resource_group_name = azurerm_resource_group.test.name
  location            = azurerm_resource_group.test.location

  administrator_login_password    = "G@Sh1DpR4!"
  coordinator_storage_quota_in_mb = 262144
  coordinator_vcore_count         = 4
  node_count                      = 2

  citus_version                        = "12.1"
  coordinator_public_ip_access_enabled = false
  ha_enabled                           = true
  coordinator_server_edition           = "MemoryOptimized"

  maintenance_window {
    day_of_week  = 1
    start_hour   = 9
    start_minute = 1
  }

  node_public_ip_access_enabled = true
  node_server_edition           = "GeneralPurpose"
  sql_version                   = "16"
  preferred_primary_zone        = 2
  node_storage_quota_in_mb      = 262144
  node_vcores                   = 4
  shards_on_coordinator_enabled = false

  tags = {
    Env = "Test2"
  }
}

leilifremont commented 12 months ago

HI @neil-yechenwei,

I double checked on my end, yes, this setting (262144) worked for me even this Monday, I verified this from the terraform plan, But it started to fail two days ago

      + node_server_edition                  = "MemoryOptimized"
      + node_storage_quota_in_mb             = 262144
      + node_vcores                          = 2

I checked the Azure portal, in the scale section, the storage dropbox of this cluster I provisioned using 262144 is blank now(meaning it's not a valid option any more), the minimum option is 512gb (refer to the screen capture below).

Also you can refer to the document of Microsoft. The minimum storage for multi node cluster is 512gb (for single node it's different, but the node_count can't be set to 1 according to your document and there is no option of "1 node" in Azure portal), and it seems this check was enforced now.