[Benchmarking Params] Need Clarity on benchmarking Parameters

Can somebody provide good documentation and difference between following options of Benchmark in C Client

I am running C Client Benchmarking.

  1. -z --threads , load generating thread count
  2. -W --event loops, defaults value it says 1, does it mean every server node in cluster will have one 1 event loop or is it on client side ?
  3. -c --asyncMaxCommands Default 50
  4. -Y --connection pool pernode.

I want to know how these values work with each other? it is creating confusion to us for benchmarking results.

The benchmark settings only apply to the benchmark client. I’m not an expert on these parameters but here’s my take on them;

  1. -z threads, for single-threaded how many theads to generate load. Ignored if specifying async params AFAIK.
  2. -W if you compiled with async, these are local (client) event loops.
  3. -c Maximum number of outstanding async commands pending locally