My ideas are:
- Multithreading using [migration_thread_pool](https://grpc.github.io/grpc/python/grpc_asyncio.html#create-server). Each thread has its own event loop and coroutines can be executed in different threads and it could utilize all CPUs. But coroutines must be thread-safe as coroutines executed in different threads can access the same data. - Multiprocessing. Each process has its main thread where event loop runs. In this case we don't need care about thread-safety as each process has its own memory space. Also read-only large data such as machine learning models can be shared in copy-on-write manner. Which is better or is there any better way? Any other considerations? -- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/grpc-io/b136e2cb-ce3d-470d-9cb0-03e67785b33dn%40googlegroups.com.
