Changes to azurerm_traffic_manager_profile destroys endpoints

[vidarw]

Community Note

• Please vote on this issue by adding a 👍 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

Terraform (and AzureRM Provider) Version

Affected Resource(s)

• azurerm_traffic_manager_profile
• azurerm_traffic_manager_endpoint

Terraform Configuration Files

# Copy-paste your Terraform configurations here - for large Terraform configs,
# please use a service like Dropbox and share a link to the ZIP file. For
# security, you can also encrypt the files using our GPG public key: https://keybase.io/hashicorp

resource "azurerm_traffic_manager_profile" "api" {
  name                = "my-api"
  resource_group_name = var.resource_group_name
  traffic_routing_method = "Priority"

  dns_config {
    relative_name = "my-api.example.com"
    ttl           = 15
  }

  monitor_config {
    protocol                     = "https"
    port                         = 443
    path                         = "/health"
    interval_in_seconds          = 30
    timeout_in_seconds           = 5
    tolerated_number_of_failures = 3
    expected_status_code_ranges  = ["100-403"]
  }
}


resource "azurerm_traffic_manager_endpoint" "api_blue_endpoint" {
  name                = "my-api-blue"
  resource_group_name = var.resource_group_name
  profile_name        = azurerm_traffic_manager_profile.api.name

  target_resource_id = var.blue_ip_resource_id
  type               = "azureEndpoints"
  priority           = 10
  endpoint_status    = var.enable_blue == true ? "Enabled" : "Disabled"

  custom_header {
    name  = "Host"
    value = "my-apu."
  }
}

Debug Output

Panic Output

Expected Behavior

Changes to values in the azurerm_traffic_manager_profile should not affect the azurerm_traffic_manager_endpoint.

Actual Behavior

Changes to values in the azurerm_traffic_manager_profile like monitor_config.path or dns_config.ttl causes all the azurerm_traffic_manager_endpoint to disappear. You basically have to run terraform twice to reprovision them.

Steps to Reproduce

1. terraform apply
2. Change monitor_config.path from /health to /healthz.
3. terraform apply

Your endpoints are now gone.

Important Factoids

References

• #0000

[djsly]

Seeing the same thing here... doing any change to the profiles wipes the endpoints...

[djsly]

@katbyte / @tombuildsstuff how we do normally proceed with this type of debugging ? From what I can gather, it seems that something changed at the MSFT side maybe ? assuming this was never an issue before ?

We just caused major downtime by changing a simple monitor setting since it wiped ALL endpoints.

I looked at the terraform code and the traffic_manager_profiles resource doesn't set any endpoints. The model can receives a list, but instead terraform uses the azurerm_traffic_manager_endpoint resource.

https://github.com/Azure/azure-sdk-for-go/blob/master/services/trafficmanager/mgmt/2018-04-01/trafficmanager/models.go#L701

So... how is this normally reported to MSFT, knowing that making changes in the portal or AZ CLI doesn't affect it seems...

[simonlegeek]

We also kicked ourselves in the nuts with this bug trying to apply a simple change in resource tags. Does anyone have an update on this?

[nitzanm]

I've created a pull request with a fix: #7471 . It was attached to a closed/duplicate issue so I'm mentioning it here to make sure it's seen.

[djsly]

@nitzanm / @ArcturusZhang which of the two PR should I follow to know when this is ready ?

@katbyte I see you are reviewing #7846 , should we expect this in 2.21.0 ? :)

[ghost]

This has been released in version 2.23.0 of the provider. Please see the Terraform documentation on provider versioning or reach out if you need any assistance upgrading. As an example:

provider "azurerm" {
    version = "~> 2.23.0"
}
# ... other configuration ...

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. If you feel I made an error 🤖 🙉 , please reach out to my human friends 👉 [email protected]. Thanks!