Container Apps: Unable to Create Environment in Central US — Capacity Error

Since September 30, 2026, every attempt to create a new Azure Container Apps managed environment in Central US has failed with:

Code: ManagedEnvironmentCapacityHeavyUsageError Message: Creating a new cluster is unavailable at this time in region centralus. To create a new cluster, we recommend using an alternate region. Inner error: AKSCapacityHeavyUsage (HTTP 400) Environment configuration:

Workload profiles environment, Consumption profile only (no dedicated profiles)

Zone redundancy: enabled

Custom VNet (dedicated /22 subnet delegated to Microsoft.App/environments)

Internal load balancer (internal-only environment)

Created with Terraform (azurerm_container_app_environment); the same error also appears when creating through the Azure portal

What we've already checked:

The same request has failed 8+ times over 5 days, through both Terraform and the portal.

The same configuration from the same Terraform module, in the same subscription and region, succeeded on September 16 and September 23.

The failures started September 30.

The target subnet is clean: no leftover service association links and no orphaned or failed environment.

Azure Service Health shows no advisory for Central US during the failure window.

Using an existing environment, or switching to managed (non-VNet) networking or a non-zone-redundant environment, isn't an option for this production workload. Central US is our primary region, so moving to another region would mean significant architectural changes.

A support case is open, and correlation IDs are available to Microsoft on request.

My questions are:

1. Is there a known capacity constraint for new Container Apps environments in Central US, especially zone-redundant ones, and is there an expected date for it to be resolved?
2. Is there a process to request a capacity exception or reservation for a specific subscription in a constrained region?
3. Has anyone else seen this recently in Central US, and did anything work, such as retrying at certain times or a different request shape?