APIGEE-API-ENGINEER practice question 22 of 25

Which statements ate true when the following policy is used? Select all that are correct. 5ps

Full set: 126 questions
Answer options
A: The rate of 5 per second indicates that 5 requests can be made at any time during a second: but the 6th and later requests dunng the same second would be rejected
B: Spike arrest is calculated separately on each message processor, so you could end up with significantly more than 5 calls in a second Correct answer
C: It is possible to make fewer than 5 requests in a second and still cause a spike arrest Correct answer
D: The continueOnError setting means that normal flow processing will resume even if the traffic exceeds the spike arrest rate
Correct Answer:
B. Spike arrest is calculated separately on each message processor, so you could end up with significantly more than 5 calls in a second; C. It is possible to make fewer than 5 requests in a second and still cause a spike arrest
Explanation

The SpikeArrest policy in Apigee is designed to throttle traffic and prevent sudden spikes, rather than acting as a strict quota. The rate of 5ps (5 per second) is translated into a minimum time interval between requests, which is 1/5 = 0.2 seconds (200ms). A request will be rejected if it arrives less than 200ms after the previous one, even if the total number of requests within that second is less than five (C). Furthermore, in a distributed environment, each Apigee Message Processor enforces the SpikeArrest policy independently. If an environment has multiple Message Processors, the total number of requests processed across all of them can exceed the configured rate, as each processor will allow up to 5 requests per second (B).

Why Incorrect

A. This describes a quota system, but SpikeArrest works by enforcing a time interval between requests, not by counting requests within a fixed one-second window.
D. The attribute continueOnError="false" explicitly instructs the proxy to stop processing the flow and enter an error state if the policy fails (i.e., the rate is exceeded).

References

1. Google Cloud

Apigee Documentation

SpikeArrest policy: "SpikeArrest differs from the Quota policy

which limits the total number of requests for a client app over a period of time... It's possible for a spike arrest to be triggered for an API that receives a low volume of overall traffic." This supports the reasoning against option A and for option C.

2. Google Cloud

Apigee Documentation

SpikeArrest policy

"How does spike arrest work?": "Think of SpikeArrest as a way to generally protect against traffic spikes rather than as a way to limit traffic to a specific number of requests... For example

a rate of 30pm means that requests are allowed roughly every 60/30 = 2 seconds." This confirms the interval-based calculation

supporting the correctness of C.

3. Google Cloud

Apigee Documentation

SpikeArrest policy

"Spike Arrest and Message Processors": "Spike Arrest policies are maintained and enforced on each Message Processor independently... Therefore

it is possible to exceed your spike arrest rate if you have more than one Message Processor." This directly supports the correctness of option B.

4. Google Cloud

Apigee Documentation

Flow variables reference

"Common policy attributes that control flow": For the continueOnError attribute: "If set to false (the default)

when a policy fails

the message is not processed further in the pipeline; it immediately enters the error flow." This directly refutes option D.

Scroll to Top