Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud Architect practice bank has 2,335 more like it. Try 10 free: instant score, no account.
A data processing application running on Google Cloud programmatically queries Google Cloud APIs to monitor thousands of compute resources. During peak operations, the application encounters transient rate-limiting and temporary service unavailability errors. Currently, the application executes immediate retries at fixed intervals, causing synchronized traffic bursts that worsen backend congestion. Furthermore, the application retrieves full resource definitions when it only requires resource state fields, causing unnecessary network overhead.
You need to optimize the application's programmatic API interactions to improve resilience and minimize network latency.
Which approach should you implement?
Exponential backoff is an error-handling strategy where the delay between consecutive retry attempts increases exponentially (for example, doubling after each failed attempt) rather than remaining constant. Jitter introduces randomized variance into these delay calculations to desynchronize simultaneous client retries. Combined with partial responses (or field masks), client applications request only the specific subset of data fields necessary for processing, discarding unnecessary payload elements at the API server level.
fields parameter or field masks) minimizes payload size over the wire, lowering serialization overhead, parsing CPU cycles, and network transfer latency.[1, RETRY_DELAY]) to smooth out aggregate traffic spikes.This architecture directly resolves both the transient communication bottlenecks and the payload efficiency issues. It aligns with Google Cloud programmatic API best practices by utilizing native client library retry configurations and request-level filtering without adding extraneous infrastructure components.
This approach places an intermediate API management layer (such as Apigee) in front of the target services and configures a ResponseCache policy to store HTTP 5xx server error responses, while relying on client-side decompression of full JSON responses.
true to avoid caching 5xx backend errors.A known static service outage of deterministic duration occurred and downstream clients required rapid synthetic maintenance responses without reaching the backend.
Response caching should be reserved for successful HTTP GET requests (200–205) with appropriate cache keys and invalidation policies.
This approach immediately reissues failed API calls without delay in a continuous while loop until a success code is returned, while fetching the entire JSON resource graph and filtering out unused attributes inside the client runtime.
maxAttempts or totalTimeoutDuration, transient server issues can cause client threads to hang indefinitely.The network connection was entirely lossless, API calls were local intra-process lookups, and the endpoint did not enforce any rate limits or quotas.
This configuration turns off all automated client-side retries, treating any transient failure or rate-limit violation as a terminal error, while using streaming protocols to receive uninterrupted feeds of complete resource models.
UNAVAILABLE or DEADLINE_EXCEEDED).The workload involved real-time binary audio/video pipelines where dropped packets are discarded rather than retransmitted and where payload structures cannot be filtered by fields.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated