Question 490
At fictional Lakeshore Finance, an HPE edge-to-cloud solution architect supporting transaction services is assessing industry architectures and deployment-model tradeoffs during a presales technical consultation (HPE-HPE0-V27-TRENDS-123). The reported customer challenge is: A business selects a data-center deployment by maximum advertised IOPS but ignores latency percentiles. Which repeatable review, test, or governance safeguard would best prevent this problem from recurring?
Explanation
Correct Answer
The best ongoing safeguard is: Require repeatable full-stack performance scenarios for purchasing decisions. It specifically reduces recurrence of: A business selects a data-center deployment by maximum advertised IOPS but ignores latency percentiles. The relevant correction is: Collect application response-time requirements and comparable I/O size and concurrency benchmarks. Proof of the corrected state is: Measured p95 and p99 storage response meets the application objective at peak demand.
Incorrect Answers
“Audit key, privileged identity and access-role lifecycle together” supports the separate task “Combine encryption keys with identity authorization and traceable data access”, not the reported problem: a business selects a data-center deployment by maximum advertised IOPS but ignores latency percentiles. This case instead calls for “Collect application response-time requirements and comparable I/O size and concurrency benchmarks”.
“Perform recovery tests against credential compromise and destructive commands” concerns the alternative task “Add independent protected recovery copies, access separation and tested restore procedures”, but the stated failure is that a business selects a data-center deployment by maximum advertised IOPS but ignores latency percentiles. The required measure is to collect application response-time requirements and comparable I/O size and concurrency benchmarks.
The proposed “Record data locations and perform cross-cloud transfer reviews at each design milestone” tests a different control: identify approved regions, subprocessors, contractual controls and allowed transfer paths before placement. The reported condition is that a business selects a data-center deployment by maximum advertised IOPS but ignores latency percentiles; the more direct action is to collect application response-time requirements and comparable I/O size and concurrency benchmarks.