For note-taking application scenarios with many reads and few writes, D1 read operations are completed at the local edge with extremely low latency, while write operations are routed to the primary node with latency typically between 50–200 ms, which is an ideal trade-off.