Which component are you using?:
Cluster-autoscaler

Is your feature request designed to solve a problem? If so describe the problem this feature should solve.:

If the checkcapacity ProvisioningRequest cannot be provisioned, it waits for backoff time before being retried. From the perspective of integration with Kueue, this time is wasteful as we block quota on Kueue's side, even it's not actively used.

Describe the solution you'd like.:
Add ProvisioningRequest Parameter that disables retry mechanism in CA. If the Parameter is present and the request cannot be provisioned it should be marked as Failed=True instead of Provisioned=False

Describe any alternative solutions you've considered.:
Add flag to the CA that disables retry mechanism. However we need more granular solution, because of different nature of some workloads - depending on a workload users are willing to block(or not) quota

Additional context.: