You have multiple instances of an API service behind a load balancer. Each client should be limited to N requests per minute, enforced consistently across all instances.
Requirements
·Rate limit counters must be shared across all service instances, not per-instance
·Checking the limit should be fast (sub-millisecond range)
·Traffic should be spread across multiple service instances
·Rate limit counters must be shared across all service instances, not per-instance
·Checking the limit should be fast (sub-millisecond range)
·Traffic should be spread across multiple service instances