Q: 5
A developer is creating a mobile app that calls a backend service by using an Amazon API Gateway
REST API. For integration testing during the development phase, the developer wants to simulate
different backend responses without invoking the backend service.
Which solution will meet these requirements with the LEAST operational overhead?
Options
Discussion
D . Mock integration with mapping templates is way less overhead than spinning up Lambda or EC2.
Mock integration is literally what API Gateway built for stuff like this. D
D , mock integration with request mapping template is exactly what you use to simulate backend responses with barely any setup. No Lambda or EC2 to manage, just config in API Gateway. Pretty sure AWS designed it for this. Disagree?
D imo. C's a bit of a trap since customizing the stage won't simulate backend responses by itself, it's mostly for deployment settings. Mock integration with mapping templates in D is the low-effort way AWS wants you to do this for testing. Always seen D picked in similar practice sets.
D , mock integration lets you return static/mock data directly from API Gateway-no Lambda or EC2 needed. Bare minimum ops effort and built for this exact testing use case. Way easier than customizing a stage or spinning up resources. Anyone see a scenario where C would work better?
Its D here, mapping template with mock integration keeps it simple, no backend, minimal setup. Not totally sure if C is ever used alone for this, but D matches what I've seen in AWS docs.
D , but I think it's a trap-sort key doesn't impact partition throughput like hash key does. B is probably correct here based on exam reports, open to corrections.
D . Both the AWS official guide and sample exams mention using mock integration with mapping templates for simulating backend responses in API Gateway, pretty sure that's the least overhead option here. If I'm missing something let me know.
Had something like this in a mock, I picked A. Lambda proxy lets you customize responses and could simulate various outputs during integration tests. I thought this was less overhead than spinning up EC2 or customizing stages. Might be off here, correct me if I missed something.
A is out, D is what I've seen in exam reports for simulating responses with minimal setup.
Be respectful. No spam.
