If the server cannot keep up with the given workload because of some bottleneck (CPU, network, disk IO), then it cannot guarantee any response times - incoming queries will be either rejected or queued in a long wait queue, which will lead to awfully big response times. This doesn't depend on the programming language or the framework the server written in.

If you want response time guarantees, make sure the server has enough free resources for processing the given workload.