Azure PostgreSQL server is stuck in 'updating' state

Since the CLI is also blocking the restart, the control plane is completely locked on Microsoft's backend. Your only remaining option is to open a support ticket in the Azure Portal so a support engineer can manually issue a backend force-cancel on the provisioning state and release the server.

Hi Igor, hitting a persistent 'Updating' hang on an Azure PostgreSQL Flexible Server after a scale-down attempt usually means the control plane operation has decoupled from the underlying compute instance, effectively locking the ARM resource. Since you’ve confirmed there are no long-running queries or active workloads blocking the operation, the platform is likely stuck waiting on an internal infrastructure timeout that can take hours to clear on its own. To force a resolution without waiting, try running an explicit restart command via the Azure CLI (az postgres flexible-server restart), as firing a direct management operation can sometimes override the stuck state. If the CLI rejects the command because the resource is locked in an updating transition, you will need to open a quick support ticket in the Azure Portal so an engineer can issue a backend force-cancel on the provisioning state and safely release your server.