The rate limit core function aims to restrict the number of requests a client can make within a specific time frame without using Redis. Here's how it works:
- The function is initialized with the rate limit (R), which specifies the maximum number of requests allowed per hour.
- For each incoming request, the function checks the timestamp of the request and compares it with the timestamps of previous requests.
- It maintains a list of recent request timestamps in memory.
- If the number of requests made within the last hour exceeds the rate limit (R), the function rejects the incoming request.
- Otherwise, the function accepts the request and updates the list of recent request timestamps.
To scale the API gateway and handle rate limiting efficiently, i using Redis as a caching solution. Here's how it works with Redis:
- A Redis Sorted Set named "requests" is used to store the timestamps of recent requests.
- Each incoming request adds its timestamp to the "requests" Sorted Set.
- Before processing a new request, the function removes all timestamps from the Sorted Set that are older than 1 hour from the current timestamp.
- The function then checks the number of remaining timestamps in the Sorted Set.
- If the count exceeds the rate limit (R), the request is rejected; otherwise, it is accepted, and the timestamp is added to the Sorted Set.
- By using Redis, the rate limiting function can efficiently handle a large number of requests and scale across multiple instances of the API gateway.
