Layer7 API Management

Just a simple question for better understanding the difference between Rate Limit Assertion and Throughput Quota assertion.

In chapter 9 of the Policy Authoring Guide where the Throughput Quota assertion is explained, most explanations are given based on request count examples. For example the "Max" field is explained as "request limit".

But in the explanation of the counter increment and decrement options, a hint says that the "by value" setting can be used to sum up response sizes to measure bandwidth usage. In that case the "Max" field must contain the allowed bandwidth regardless of request counts.

So is it correct, that the Throughput Quota assertion is not bound to requests? It is basically an abstract quota assertion that counts anything against a max value?

If yes, this is a very important difference to the Rate Limit assertion. 

Thanks

Stephan

The Apply Throughput Quota is not bound to requests. It is bound to time. For example, "per hour" starts at the the top of the hour, not the first count of a request. MAX refers to the number of requests within the chosen time period starting at the beginning of the second/minute/hour/day/month.

The counter does apply to requests, and by default the counter increments by one for each request. However, you can override this behaviour by using the By Value field. Using By Value, you can now have each request count as 2 or 5...or use the ${response.size} variable in which case you are now counting by used bandwidth. In this case, yes, your units of measure are no longer in requests, but in bandwidth, and you'd have to adjust the MAX value to account for this.

For more information, see Apply Throughput Quota Assertion - CA API Gateway - 9.0 - CA Technologies Documentation