What was the load on the services? What optimizations were applied?

The load on services was usually measured in requests per second (RPS), the volume of data processed, and response time. For example, a service could handle up to several thousand requests per second with a latency of less than 100 ms.

The following approaches were used for optimization:

• Caching at different levels (in-memory, Redis) to reduce database load.
• Using asynchronous calls and Kotlin coroutines for efficient thread management.
• Optimizing database queries, adding indexes.
• Horizontal scaling of services through containerization and orchestration.
• Profiling and eliminating bottlenecks in the code, such as excessive synchronization or locking.

Example of using coroutines for asynchronous processing:

suspend fun fetchData(): Data {
    return withContext(Dispatchers.IO) {
        // request to database or external API
    }
}