Set memory limits (#921)

We're seeing OOM issues at the Node level when we don't set the limit on some pods. This will allow the pod to restart gracefully w/o taking down the entire node.

I'm not stuck on the specific value so we can pick a different one.

See knative/serving#3332 for the serving equivalent.

Signed-off-by: Doug Davis <dug@us.ibm.com>

config/500-controller.yaml

@@ -52,6 +52,9 @@ spec:
         ports:
           - containerPort: 9090
             name: metrics
+        resources:
+          limits:
+            memory: 1000Mi  # An initial guess, but consistent with serving
         volumeMounts:
           - name: config-logging
             mountPath: /etc/config-logging

config/500-webhook.yaml

@@ -46,6 +46,9 @@ spec:
               fieldPath: metadata.namespace
         - name: CONFIG_LOGGING_NAME
           value: config-logging
+        resources:
+          limits:
+            memory: 1000Mi   # An initial guess, but consistent with serving
       volumes:
         - name: config-logging
           configMap: