Enable worker to handle a large number of client requests

Currently, suppose the number of incoming requests is much greater than the number of worker threads in a RS worker, many of the requests will fail because the server is busy. The client will retry, however they may still fail because the requests are too many.

The above scenario looks pretty common because RS is a shared component between the computing and storage layers. I would also assume it is common that a RS worker will receive many client requests at the same time. Therefore, we need to have a better solution for this issue.