What is the reason the genesyscloud_byoc_private_edge resource fails to allocate the next available IP address in the defined range when using Terraform provider v1.9.4?
Deploying a new BYOC private edge instance in the au-1 region. The infrastructure is managed via GitHub Actions. The deployment succeeds on the first run, creating the edge and assigning 10.0.0.51. Subsequent runs that add a second edge instance fail with a state lock or a 409 Conflict error from the API, claiming the IP is already in use, even though the previous apply completed successfully and the state file is updated.
The HCL configuration uses a dynamic block for the IP assignment. I have verified the CIDR block 10.0.0.0/24 has sufficient capacity. The issue seems to stem from the asynchronous nature of the IP allocation endpoint in the Genesys Cloud API. The Terraform provider appears to mark the resource as created before the IP is fully propagated across the control plane, leading to race conditions when multiple edges are defined in the same module.
Here is the relevant snippet:
resource "genesyscloud_byoc_private_edge" "edge_node" {
name = "byoc-edge-${count.index}"
description = "Private edge for BYOC deployment"
ip_address = "10.0.0.${50 + count.index}"
network {
gateway = "10.0.0.1"
netmask = "255.255.255.0"
}
}
Adding a time_sleep dependency before the apply helps temporarily, but this is not a scalable solution for larger deployments. The logs show the API returns a 201 Created, but the subsequent read operation fails to fetch the IP details immediately. Is there a documented retry logic or a specific provider argument to handle this latency?
How can we ensure the Terraform provider waits for the IP allocation to fully propagate in the au-1 region before proceeding with the next resource creation, avoiding 409 conflicts during bulk BYOC edge deployments?