Hi. The use of Redis in this case does not have such an effect. It has a significantly greater effect when Redis is local, but the application connects to a database on another server. Even if the server is on the same network, the difference from localhost can be 10–100x worse due to latency (for example direct socket access or 0.01 ms vs. 0.3 ms). In this case, Redis does help by 1000–10000%. In addition, a very trivial SELECT is used in the demo, which will always be fast and return only 1 record. But in a real application, the database must also calculate, filter, sort, group, and so on.