Which technique is commonly used to mitigate I/O latency and improve OS performance, such as batching I/O, asynchronous I/O, prefetching, and caching?

• A. Disabling caching entirely.
• B. Increasing page fault threshold.
• C. Synchronous I/O only.
• D. I/O batching, asynchronous I/O, prefetching, and caching.

Answer: (B)

Mitigating I/O latency and boosting OS performance comes from overlapping work with I/O and keeping data ready when needed. Batching I/O reduces per-request overhead and disk seeks by combining multiple requests into fewer operations, which improves throughput. Asynchronous I/O lets the system start I/O and continue with other tasks, so the CPU isn’t tied up waiting for I/O to finish. Prefetching anticipates future data needs and brings that data into memory before it’s actually requested, hiding latency. Caching stores frequently used data in fast storage (RAM) so repeated accesses don’t incur slow disk I/O.

These techniques work together to reduce the time applications wait for I/O and to make better use of CPU time. In contrast, turning off caching removes a key fast path, increasing latency; increasing the page fault threshold targets memory pressure rather than directly improving I/O timing; and relying solely on synchronous I/O forces the CPU to stall on I/O, offering no overlap or latency hiding.