After speaking to Dennis, it looks like this is a Provider issue.
Moving to post, has the same fix as: https://bugzilla.redhat.com/show_bug.cgi?id=1568420 https://github.com/ManageIQ/manageiq-providers-kubernetes/pull/249
Created attachment 1458024 [details] hawkular_force_legacy Verified on 5.10.0.3.20180705151325_ce4d2b5. Please pay attention that on CF 4.7\5.10 the parameter hawkular_force_legacy default value is true, which means CF doing the metrics process, Unlike on CF 4.6\5.9 wich, the parameter hawkular_force_legacy default value was false. tested it with parameter hawkular_force_legacy value set to true & false and the CPU load of hawkular pod is around 0.2 cores.