As documented here , the endpoint /webapi-init-check can be called in order to check the health and readiness of an API GW.

I would like to know how frequently this endpoint can be called. If each call to it will trigger a bunch of internal calls to check the health of different components, then I can imagine that it should not be called too frequently, so as to avoid the API GW spending 50% of its CPU resources on running health checks. But if it is merely reading a property which is updated internally at a fixed interval, then I suppose it doesn't matter how frequently it is called.

The background for this query is that I have an API consumer who would like to check the health of the API GW so as to be able to better determine where the problem lies, in case there is a problem. And I would like to understand how this would scale if more API consumers want to do this. And also in order to understand what interval I should tell them to set between each call.

Hi Daniel,

Option 1:
You can implement a keep-alive operation in every API exposed to consumers, allowing them to invoke it based on their own flexibility.

Option 2:
Alternatively, you can create a platform-level keep-alive API, deploy it across all catalogs, and have consumers call this centralized API.This can be even called after upgrade 

Option 3:

Since the IBM-provided endpoint can be configured in F5, please set up an email notification for the operations team in case a status code other than 200 is returned