An ecommerce company experiences a surge in mobile application traffic every Monday at 8 AM during the company's weekly sales events. The…

The core issue is that a synchronous invocation pattern (API Gateway -> Lambda) cannot handle the sudden burst of traffic, leading to Lambda throttling and TooManyRequestsException errors. The most effective solution is to decouple the components using a message queue.

Amazon SQS is designed for this purpose. By placing an SQS queue between API Gateway and the Lambda function, the queue acts as a buffer, absorbing the traffic spike. The Lambda function can then pull messages from the queue and process them at a sustainable rate, preventing throttling. Using provisioned concurrency ensures that a specified number of Lambda execution environments are pre-warmed and ready, minimizing cold-start latency and further improving responsiveness for the requests processed from the queue.

B: AWS Step Functions is an orchestration service for multi-step workflows. It is not designed for buffering high-volume, short-lived requests and would add unnecessary complexity and latency.

C: Amazon SNS is a pub/sub service designed for fanning out messages to multiple subscribers. It does not buffer requests; it pushes them to subscribers immediately, which would still result in a traffic spike at the Lambda function and cause throttling.

D: Reserved concurrency limits the maximum number of concurrent executions for a function. Applying a limit is counterproductive when the goal is to handle a traffic surge, as it would restrict the function's ability to scale up.

1. AWS Lambda Developer Guide

"Using AWS Lambda with Amazon SQS": This document explicitly states

"You can use an Amazon SQS queue to decouple services and create a more resilient and scalable architecture... A queue can act as a buffer between a web front end and a back-end process. When you use a queue

the front-end service can add requests to the queue and immediately respond to the user. The back-end service can then process the requests at its own pace." (Section: "Using AWS Lambda with Amazon SQS"

Introduction).

2. AWS Lambda Developer Guide

"Configuring provisioned concurrency": "Provisioned concurrency initializes a requested number of execution environments so that they are prepared to respond immediately to your function's invocations... Configuring provisioned concurrency for your function incurs charges to your AWS account." This explains how provisioned concurrency reduces latency

which complements the SQS buffering strategy. (Section: "Managing Lambda function concurrency"

Subsection: "Provisioned concurrency").

3. AWS Lambda Developer Guide

"Managing Lambda function concurrency": "To prevent a function from using too much concurrency... you can configure a function with reserved concurrency. When a function has reserved concurrency

it can't use more concurrency than the reserved amount." This confirms that reserved concurrency is a limiting mechanism

making it unsuitable for handling traffic surges. (Section: "Managing Lambda function concurrency"

Subsection: "Reserved concurrency").

4. AWS Well-Architected Framework

Reliability Pillar Whitepaper (page 23): "Decouple components: Components that are tightly coupled can impact the availability of a system... Decoupling components with queues

stream

or load balancers can help isolate component behavior." This document endorses the architectural pattern of using queues to improve reliability and handle load variations.