Negative Caching for Provider Request Lock Cache

[akashsinghal]

Describe the solution you'd like The new request provider cache locks per repository/digest combination. It caches any result returned during the verification process, either accept, deny, or error. If the ttl is set too high, this can cause problems with retries due to initial failures of pod creation (e.g. network error, registry responded with transient error). We should enforce a small ttl for this cache and consider a mechanism for users to externally invalidate this cache.