removing endpoints for controller and scheduler

... After upgrading to k3s 1.19 (from 1.18), the promethgeus target scraping for those two targets stopped wokring,

`Get "http://10.2.0.30:10252/metrics": dial tcp 10.2.0.30:10252: connect: connection refused`

Looking at k3s-io/k3s#425 (comment) suggests the endpoint approach should work.  Experimenting with removing the explicit endpoint callout to see if there is an improvement.

Signed-off-by: Jeff Billimek <jeff@billimek.com>

monitoring/kube-prometheus-stack/kube-prometheus-stack.yaml

@@ -320,12 +320,12 @@ spec:
       enabled: false
     kubeControllerManager:
       enabled: true
-      endpoints:
-      - 10.2.0.30
+      # endpoints:
+      # - 10.2.0.30
     kubeScheduler:
       enabled: true
-      endpoints:
-      - 10.2.0.30
+      # endpoints:
+      # - 10.2.0.30
     kubeProxy:
       enabled: false
     prometheus-node-exporter: