Deployment failed with subnet state not ready to accept private endpoint (possible dependency mismatch?)

[SvenAelterman]

Describe the bug

Doing a deployment with make all, I received the following error message:

Error: waiting for creation of Private Endpoint "pe-blob-actredemo01" (Resource Group "rg-actredemo01"): Code="RetryableError" Message="A retryable error occurred." Details=[{"code":"ReferencedResourceNotProvisioned","message":"Cannot proceed with operation because resource /subscriptions/<snip>/resourceGroups/rg-actredemo01/providers/Microsoft.Network/virtualNetworks/vnet-actredemo01/subnets/SharedSubnet used by resource /subscriptions/<snip>/resourceGroups/rg-actredemo01/providers/Microsoft.Network/networkInterfaces/pe-blob-actredemo01.nic.4b12ee8f-5b4a-4d42-adfc-d891f82c99f7 is not in Succeeded state. Resource is in Updating state and the last operation that updated/is updating the resource is PutSubnetOperation."}]
│ 
│   with azurerm_private_endpoint.blobpe,
│   on storage.tf line 18, in resource "azurerm_private_endpoint" "blobpe":
│   18: resource "azurerm_private_endpoint" "blobpe" {
│

It would seem to indicate that the deployment of a private endpoint is attempted before the network/subnet has finished creating.

When I redeployed, I got an error that the private endpoint resource already existed and should be imported. Instead, I chose to delete the resource and re-ran the deployment again. It then finished successfully.

Steps to reproduce

1. Following the deployment steps of the Quickstart guide, for a greenfield deployment.

[marrobi]

I've just had similar this morning:

╷
│ Error: waiting for creation of Private Endpoint "pe-file-mrtredemo27" (Resource Group "rg-mrtredemo27"): Code="RetryableError" Message="A retryable error occurred." Details=[{"code":"ReferencedResourceNotProvisioned","message":"Cannot proceed with operation because resource /subscriptions/xxx/resourceGroups/rg-mrtredemo27/providers/Microsoft.Network/virtualNetworks/vnet-mrtredemo27/subnets/SharedSubnet used by resource /subscriptions/xxx/resourceGroups/rg-mrtredemo27/providers/Microsoft.Network/networkInterfaces/pe-file-mrtredemo27.nic.252711c3-56f3-462e-adbe-a4e10e229764 is not in Succeeded state. Resource is in Updating state and the last operation that updated/is updating the resource is PutSubnetOperation."}]
│ 
│   with azurerm_private_endpoint.filepe,
│   on storage.tf line 39, in resource "azurerm_private_endpoint" "filepe":
│   39: resource "azurerm_private_endpoint" "filepe" {
│ 
╵

╷
│ Error: A resource with the ID "/subscriptions/xxx/resourceGroups/rg-mrtredemo27/providers/Microsoft.Network/privateEndpoints/pe-file-mrtredemo27" already exists - to be managed via Terraform this resource needs to be imported into the State. Please see the resource documentation for "azurerm_private_endpoint" for more information.
│ 
│   with azurerm_private_endpoint.filepe,
│   on storage.tf line 39, in resource "azurerm_private_endpoint" "filepe":
│   39: resource "azurerm_private_endpoint" "filepe" {
│ 
╵

[marrobi]

Similar to what is discussed here:

hashicorp/terraform-provider-azurerm#16182

Are a few other references, I think we should maybe make all the private endpoint operations happen sequentially?

Would be useful to have the full log from core/terraform/*-tre-core.log if anybody can provide one.

[marrobi]

Also seen:

Cannot proceed with operation because resource /subscriptions//resourceGroups/rg-mrtredemo27-ws-9f98/providers/Microsoft.Network/virtualNetworks/vnet-mrtredemo27-ws-9f98 used by resource /subscriptions//resourceGroups/rg-mrtredemo27-ws-9f98/providers/Microsoft.Network/virtualNetworks/vnet-mrtredemo27-ws-9f98/virtualNetworkPeerings/ws-core-peer-mrtredemo27-ws-9f98 is not in Succeeded state. Resource is in Updating state and the last operation that updated/is updating the resource is PutSubnetOperation.

When deploying a workspace occasionally get the above error. Peering and private endpoint operations need to wait until the subnets are deployed.