Option D is correct because a 10% cache hit ratio means 90% of data requests are going to disk, which is extremely poor performance - significantly increasing cache size will allow more data to stay in memory, reducing expensive disk operations. Changing application code (A) might help but cache sizing is the direct fix. Running queries twice (C) is nonsensical, and doing nothing (B) with a 10% hit ratio is unacceptable - this indicates severe cache starvation.