I am investigating one issue related to the rate limiting in my server. Previous developer has setup spike arrest and quota policy in ApiGee. I read the documentation but I am unable to understand how both the policy works parallelly?
For example:
Client (Web, Mobile) calling the API. There are more than 100 concurrent users access the API. So which policy applied? Spike arrest or Quota?
If anyone has real world idea about this then please provide some insight.
Thanks
The particular behavior of the API proxy will depend on the placement of the two policies within flows, but assuming a standard request flow with serial policies, then generally the spike-arrest policy will protect your back-end services in aggregate, while the quota policy will enforce rate-limits on some chosen client-specific criteria. Thus one is a general overall safety protection for your business-logic back-end (spike arrest), and the other is more for enforcing client-specific constraints as dictated by your end-to-end application design and expected use-case interactions (quota). Both are configurable though, so the details of those configurations matter in the final analysis.
Comparison docs are here: https://docs.apigee.com/api-platform/develop/comparing-quota-spike-arrest-and-concurrent-rate-limit-policies
