How is the asynchronous service related to a 25% reduction in load on PostgreSQL?

An asynchronous service allows to offload the PostgreSQL database by changing the request processing model:

• Instead of performing long or resource-intensive operations synchronously and blocking database connections, the service queues tasks and processes them asynchronously.
• This reduces the number of concurrent requests to the database, decreases locks and resource contention.
• As a result, the load on PostgreSQL decreases, which can lead to reduced CPU, memory usage, and response time.

For example, if previously the application directly wrote data to the database with each user request, now it places tasks in a queue (e.g., RabbitMQ, Kafka), and a separate worker processes these tasks asynchronously and updates the database.

This approach can reduce peak load on PostgreSQL by about 25% and increase overall system performance.