What was the RPS on your services at Wildberries?

In Wildberries, RPS (requests per second) on services varied greatly depending on the type of service and time of day. For example:

• Main user API services handled from a few thousand to tens of thousands of requests per second.
• Services related to catalog and search could reach hundreds of thousands of RPS during peak hours.
• Internal services, such as order processing or logistics, had a more moderate load — hundreds or thousands of RPS.

To support such loads, horizontal scaling, caching, efficient algorithms, and database optimization were used.

Specific figures depend on the period, architecture, and tasks, but overall, Wildberries services are designed for high load with peak values in the tens and hundreds of thousands of RPS.