Question 1
Q1A storage architect is investigating intermittent high latency on an HPE Alletra 9000 array that supports a critical OLTP database. Using HPE InfoSight, the architect observes that while write latency remains consistently under 1ms, read latency occasionally spikes to 15ms during batch reporting jobs. What is the most likely root cause of this behavior?
Show answer & explanation
Correct answer: B
In storage arrays like the HPE Alletra 9000, writes are immediately acknowledged once they hit the non-volatile cache (resulting in low write latency). However, large batch reporting jobs typically generate massive sequential reads. If these reads are not already cached, they result in cache misses, forcing the array to fetch data directly from the backend NVMe drives. This sudden influx of backend I/O causes read latency spikes.