[INTERNAL] - Broker CPU utilization underestimated on Kubernetes

See: linkedin#1242
This is an internal patch to use SystemCpuLoad to monitor broker pod
cpu usage knowing that we always run the broker in a container.

As `SystemCpuLoad` reports un-normalized cpu load across all cores
we do this normalization to match CC expected value in [0, 1] interval.

cruise-control-metrics-reporter/src/main/java/com/linkedin/kafka/cruisecontrol/metricsreporter/metric/MetricsUtils.java

@@ -403,10 +403,15 @@ private static CruiseControlMetric toCruiseControlMetric(long nowMs,
    * @return the "recent CPU usage" for the JVM process as a double in [0.0,1.0].
    */
   public static BrokerMetric getCpuMetric(long nowMs, int brokerId, boolean kubernetesMode) throws IOException {
-    double cpuUtil = ((com.sun.management.OperatingSystemMXBean) ManagementFactory.getOperatingSystemMXBean()).getProcessCpuLoad();
-
+    double cpuUtil;
     if (kubernetesMode) {
-      cpuUtil = ContainerMetricUtils.getContainerProcessCpuLoad(cpuUtil);
+      // we are using openjdk>=14 runtime in k8s environments so
+      // the process cpu utilization is computed as: system_load/no_of_processors
+      com.sun.management.OperatingSystemMXBean systemMXBean =
+              (com.sun.management.OperatingSystemMXBean) ManagementFactory.getOperatingSystemMXBean();
+      cpuUtil = systemMXBean.getSystemCpuLoad() / systemMXBean.getAvailableProcessors();
+    } else {
+      cpuUtil = ((com.sun.management.OperatingSystemMXBean) ManagementFactory.getOperatingSystemMXBean()).getProcessCpuLoad();
     }
 
     if (cpuUtil < 0) {