
Configuring a rate, burst, or count limit on the DataPower API Gateway
If you want to include a rate limiting policy in your API assembly flow, you must first configure the required rate, burst, and count limits on the Gateway. Rate and burst limits restrict the number of calls that an application can make to API in specified time period, while count limits impose a strict limit on the total number of calls.
About this task
You configure rate, burst, and count limits on the DataPower® API Gateway by defining them in one or more configuration files that you package and then add to the Gateway as a Gateway extension.
You define Catalog scoped rate, burst, and count limits in the appropriate
api-collection
object on the DataPower API Gateway, which is the
object that represents your API Connect Catalog in the gateway
configuration. The defined rate, burst, and count limits will be available to be included in Rate
Limit policies in the assembly flows of all the APIs that are published to that Catalog.
You define Gateway scoped rate, burst, and count limits in
the appropriate apigw
object on the DataPower API Gateway. The defined rate,
burst, and count limits will be available to be included in Rate
Limit policies in the assembly flows of all the APIs that are published to that Gateway.
- 1 second interval starts at each new second.
- 1 minute interval starts at the beginning of each minute.
- 1 hour interval starts at the beginning of each hour.
- 1 day interval starts at midnight UTC.
- 1 week interval starts on Thursday at 00:00:00 UTC.
- Only one Gateway extension can be in use at any one time, so you should package all required rate, burst, and count limits in a single Gateway extension.
- To apply a Catalog scoped or Gateway scoped rate, burst, or count limit, it must be referenced in a Rate Limit policy in an API assembly.
- You should ensure that all rate, burst, and count limit names are unique across all your Plans and Catalogs. When an API is called, information about any rate, burst, or count limits (such as name, limiting criteria, and remaining number of calls) is returned in the HTTP headers of the response. If two or more limit definitions with exactly the same name are applied during the invocation of an API, information about only one of them is returned. However, all limiting criteria are applied.