You have reached the quota for a model. How can you create another deployment of it?
Why A is the answer
Reduce the TPM assigned to other deployments of that model
Microsoft says that once you reach the limit, you must reduce TPM on other deployments, or request more quota.
In detail
- TPM is shared across deployments of a model.
- Reducing one frees quota for another.
- Review idle deployments regularly.
Use case
An illustrative scenario, not a real incident.
A team is out of quota for a model, but its test deployment holds spare TPM. Lowering it frees room for the new deployment.
See it in different ways
flowchart LR A["Subscription quota per region, model and type"] --> B["Deployment A: TPM"] A --> C["Deployment B: TPM"] B --> D["Requests over the limit get throttled"]
stateDiagram-v2 [*] --> ModelQuota ModelQuota --> DeploymentA: assign TPM ModelQuota --> DeploymentB: assign TPM DeploymentA --> ModelQuota: reduce TPM to free capacity DeploymentB --> Throttled: peak above its limit
flowchart LR S0["Reduce TPM on one"] --> S1["Quota freed"] --> S2["Create new deployment"]
Key point: Free TPM by reducing another deployment.
Why the other options are wrong
- B. Move the resource into a different folder of the same project
- Folders have no effect on quota.
- C. Restart each deployment so that the quota counter resets
- Restarting does not reset quota.
- D. Rename the existing deployment so that it is counted as new
- Renaming changes nothing.