You can control your application's throughput consumption by configuring throughput buckets in a container and specifying a maximum throughput percentage for each one. When a bucket exceeds its assigned limit, incoming requests are throttled to maintain isolation and avoid contention between different applications sharing the same container.