We want to normalize CPU utilization (0-100%) as dashboards do not show core count and NATS servers may run on different hardware/instance types. - [ ] Add another metric that normalizes the values - [ ] Update dashboards to use the normalized metric