Performance Why don't I get the throughput I benchmarked? In synthetic benchmarks, you can achieve a high throughput for your system by throwing lots of independant tasks/clients at it to see what the theoretical limit of your system is, however in
Benchmarking Latency for a set Throughput Without a set throughput, a latency measurement might be meaningless. Often you see benchmark results with throughput measurements, and sometimes the include latency measurements. However, latency varies with the throughput you pe