Provider produced inconsistent result after apply

[eshirke]

how to handle below error for landscape_label, below are the error from terraform.

Error: Provider produced inconsistent result after apply
When applying changes to module.global-bas-nonprod.btp_subaccount_environment_instance.cloudfoundry, provider "provider["registry.terraform.io/sap/btp"]" produced an unexpected new value: .landscape_label: was cty.StringVal("us30"), but now cty.StringVal("cf-us30").

This is a bug in the provider, which should be reported in the provider's own issue tracker.

[lechnerc77]

Can you share your configuration. First guess is that you set an invalid value for the landscape label in your configuration that was overwritten by the computed value which could cause this error.

[eshirke]

I have created module for subaccount and its related configuration. I am using region variable across all configurations like below. seems like cloud foundry has cf- naming stanadards as per sap documentation (https://help.sap.com/docs/btp/sap-business-technology-platform/regions-and-api-endpoints-available-for-cloud-foundry-environment).

module "global-bas-nonprod" {
source = "./modules/subaccount"
environment = "nonprod"
business_unit = "global"
btp_service = "bas"
region = "us30"
subaccount_users = local.bas_subaccount_roles
entitlements = {
"sapappstudio" = ["standard-edition"],
}
}

[lechnerc77]

It would be of interest what parameters you use as parameter for the resource btp_subaccount_environment_instance(probably inside the module).
From what you are explain you are using the region as landscape label. The landscape label is equal to the technical key in the documentation you referenced, so prefixing the variable region with cf- and using this as input for the resource btp_subaccount_environment_instance should do the trick in your scenario.

[eshirke]

Yes, for now, I’m using the approach you suggested. Surprisingly, in the earlier case, Terraform didn’t show any errors during terraform plan even without the "cf-" prefix.

resource "btp_subaccount_environment_instance" "cloudfoundry" {
subaccount_id = btp_subaccount.subaccount.id
name = btp_subaccount.subaccount.name
environment_type = "cloudfoundry"
service_name = "cloudfoundry"
plan_name = "standard"
landscape_label = "cf-${var.region}"
parameters = jsonencode({
instance_name = btp_subaccount.subaccount.name
})
}

[lechnerc77]

The reason for that is that the terraform plan does not apply the change i.e., does not call the API that creates the environment instance. In your case the check of the state (planned vs. current state based on API response) that Terraform executes in terraform apply namly after the provisioning was successful leads to the mismatch and the error shown above.

It would be better if the API would raise an error when you provide an invalid value, but obviously it just ignores it.