PERF(grafana): optimize cpu requests based on actu

- code-server: 50m → 5m (actual: 1m)
- docusaurus: 50m → 5m (actual: 1m)
- kubernetes-dashboard: 15m → 5m (actual: 1m)
- kubernetes-dashboard-metrics-scraper: 50m → 5m (actual: 1m)
- homer: 50m → 5m (actual: 1m)
- umami: 30m → 5m (actual: 1m)
- gitea: 50m → 10m (actual: 5m)

homer/homer.yaml

@@ -65,7 +65,7 @@ spec:
         resources:
           requests:
             memory: "64Mi"
-            cpu: "50m"
+            cpu: "5m"  # Reduced from 50m based on actual usage (1m)
       volumes:
       - name: config
         configMap: