Instead of (at least as an option, but I think preferably exclusively) use libbeat monitoring output for monitoring apm-server status (activity/quiescence), and for producing reports. This would enable running hey-apm against server deployments where exvpar is not or can not be enabled, such as ESS.
Changed my mind, not worth pursuing for the time being. I'd rather focus on spend effort on improving tooling for comparing benchmarks, identifying and isolating the cause of deltas.
Instead of (at least as an option, but I think preferably exclusively) use libbeat monitoring output for monitoring apm-server status (activity/quiescence), and for producing reports. This would enable running hey-apm against server deployments where exvpar is not or can not be enabled, such as ESS.