I haven't used grpc, but benchmarks are notoriously difficult to get right and often end up measuring the wrong thing or being misleading. It does seem a little odd that the rps actually goes up with more concurrent clients.
I haven't used grpc, but benchmarks are notoriously difficult to get right and often end up measuring the wrong thing or being misleading. It does seem a little odd that the rps actually goes up with more concurrent clients.