We will go for the prometheus metrics approach, we will have a pod (or multiple pods) deployed in the cluster and retrieving metrics from the nodes accessing their prometheus metrics endpoint. At the end of (or during) the test this pod(s) will expose the results on a WebSockets endpoint. These pods will implement the results service as outlined here https://github.com/w3f/polkadot-lab/wiki/Intro#results-service
We will go for the prometheus metrics approach, we will have a pod (or multiple pods) deployed in the cluster and retrieving metrics from the nodes accessing their prometheus metrics endpoint. At the end of (or during) the test this pod(s) will expose the results on a WebSockets endpoint. These pods will implement the results service as outlined here https://github.com/w3f/polkadot-lab/wiki/Intro#results-service