Quota Limit not precise on edge eval

[jasonwang090311]

Hi,

Apologize in advance if the format is not correct, it's my first time asking on community.

I was developing a quota limit policy based on

https://cloud.google.com/apigee/docs/api-platform/reference/policies/quota-policy#intro

However after setting up I found out that the quota limit policy was not blocking traffic perfectly. I have it set for 2 request per min. But I was able to make over 5 request before I was blocked. And a few seconds later I was able to run the request again. Is there an explanation on why this happens? Also, if I actually need to control my quota (2 per min for ex), which policy should I use to make sure it really blocks the traffic when exceeded. Thank you!

[dknezic · Accepted solution]

For your use case, ensure you have configured the policy as distributed and synchronous. This will implement a central counter

Otherwise, as the component that processes requests and executes policies - the message processor, will each have an independent counter, and there are multiple message processors for high availability and as traffic grows, etc.

https://docs.apigee.com/api-platform/reference/policies/quota-policy#distributed

[jasonwang090311]

Thank you very much for the explanation

[#4]

I have just responded to same question in another one.

ref:

https://community.apigee.com/questions/91768/quota-policy-request-are-accepted-even-after-the-q.html...