bindata: CPU requests overstate actual needs

In production environments etcd is over requested for actual use. The bulk of that over-requesting comes from the sidecars which overstate their actual needs. Adjust them based on the calculations from prod clusters to request roughly 100m less CPU than before.
openshift · Jan 27, 2021 · 7952931 · 7952931
1 parent d951824
commit 7952931
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 8 deletions.
diff --git a/bindata/etcd/pod.yaml b/bindata/etcd/pod.yaml
@@ -42,7 +42,7 @@ spec:
       resources:
         requests:
           memory: 60Mi
-          cpu: 30m
+          cpu: 10m
       securityContext:
         privileged: true
       env:
@@ -68,7 +68,7 @@ ${COMPUTED_ENV_VARS}
       resources:
         requests:
           memory: 60Mi
-          cpu: 30m
+          cpu: 10m
       securityContext:
         privileged: true
       volumeMounts:
@@ -92,7 +92,7 @@ ${COMPUTED_ENV_VARS}
     resources:
       requests:
         memory: 60Mi
-        cpu: 30m
+        cpu: 10m
     volumeMounts:
       - mountPath: /etc/kubernetes/manifests
         name: static-pod-dir
@@ -209,7 +209,7 @@ ${COMPUTED_ENV_VARS}
     resources:
       requests:
         memory: 200Mi
-        cpu: 100m
+        cpu: 40m
     securityContext:
       privileged: true
     volumeMounts:

diff --git a/pkg/operator/etcd_assets/bindata.go b/pkg/operator/etcd_assets/bindata.go