Evaluate an Application Design Considering Latency and Rate Limiting 350-901 Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A mobile app developer is integrating with a third-party API that returns HTTP 429 (Too Many Requests) responses with a Retry-After header. The app needs to handle these responses gracefully without overwhelming the API. Which client-side strategy is most appropriate?

   • AImmediately retry the request up to 10 times with no delay to ensure the request eventually succeeds.
   • BIgnore the 429 response and display an error to the user, asking them to try again later.
   • CImplement exponential backoff with jitter, respecting the Retry-After header when present.
   • DIncrease the timeout of the HTTP client to 60 seconds to give the server more time to process the request.
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   When designing an application, what is a common trade-off between latency and rate limiting?

   • ARate limiting always reduces latency by reducing server load
   • BRate limiting has no impact on latency
   • CAggressive rate limiting can increase latency for legitimate users due to queuing or retries
   • DLower latency always requires disabling rate limiting
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   Which design approach best balances the need for low latency and rate limiting to optimize user experience?

   • AImplementing client-side throttling with exponential backoff and retries
   • BUsing a fixed window algorithm with a large window size
   • CDisabling rate limiting to ensure lowest latency
   • DSetting a very low rate limit to guarantee system stability
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   A developer is building a client application that calls a third-party API with a documented limit of 10 requests per second. The client needs to make a burst of 20 requests immediately, then continue making requests at a steady rate. Which client-side rate limiting approach allows the burst while still respecting the overall limit?

   • AUse a token bucket algorithm with a bucket capacity of 10 tokens and a refill rate of 10 tokens per second.
   • BImplement a simple counter that resets every second and rejects any request after the 10th in that second.
   • CUse a leaky bucket algorithm with a bucket capacity of 20 and a leak rate of 10 requests per second.
   • DUse a fixed window counter that allows a maximum of 10 requests per 1-second window.
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A team is designing a notification service that sends emails and push notifications. The service calls a third-party email provider that has a strict rate limit of 50 requests per second. The service must handle a spike of 500 notifications when a popular item is released. The team wants to minimize the time it takes to deliver all notifications while respecting the provider's limit. Which approach best balances these constraints?

   • AIncrease the timeout on the HTTP client to 60 seconds so the provider has time to process the burst.
   • BUse a queue to buffer all 500 notifications and have a worker process consume from the queue at a rate of 50 per second.
   • CSend all 500 notifications concurrently in a single burst, relying on the provider to queue excess requests.
   • DImplement client-side throttling in the notification service to send 50 requests per second, blocking the main thread until each batch is sent.
   Reveal Solution Discussion Select an answer first