Hide Forgot
Description of problem: In web console, when editing request/limit values for dc and standalone rc that have multiple containers, it should check the requests/limits sum of all containers < Pod limit max, but it does not. Version-Release number of selected component (if applicable): fork_ami_openshift3_userinterface_public_464_200 How reproducible: Always Steps to Reproduce: 1. oc login, create a project and LimitRange in it $ cat > limits.yaml <<EOF apiVersion: v1 kind: LimitRange metadata: name: limits spec: limits: - max: cpu: 500m memory: 750Mi min: cpu: 10m memory: 5Mi type: Pod - default: cpu: 130m memory: 120Mi defaultRequest: cpu: 110m memory: 100Mi maxLimitRequestRatio: cpu: 10 memory: 8 max: cpu: 400m memory: 750Mi min: cpu: 10m memory: 5Mi type: Container EOF $ oc create -f limits.yaml -n <project> --config=./admin.kubeconfig 2. Create a standalone rc that has multiple containers $ cat > rc_containers.yaml << EOF apiVersion: v1 kind: ReplicationController metadata: name: myrc spec: replicas: 1 selector: run: myrc template: metadata: labels: run: myrc spec: containers: - image: yapei/hello-openshift imagePullPolicy: IfNotPresent name: hello-openshift ports: - containerPort: 8080 protocol: TCP - image: yapei/hello-openshift-fedora imagePullPolicy: IfNotPresent name: hello-openshift-fedora ports: - containerPort: 8081 protocol: TCP EOF $ oc create -f rc_containers.yaml 3. In web console, browse Deployments --> Click the standalone rc --> Click edit button in the right --> Set Resource Limits --> Set all containers' CPU Request=260m, Limit=300m, so that 260 + 260 > 500 Actual results: 3. The "Save" button can be clicked. Expected results: 3. Because requests/limits sum of all containers > Pod limit max, the "Save" button should be gray and prohibited. Additional info:
Fix included in https://github.com/openshift/origin/pull/6853
Verified against devenv-rhel7_3421 (openshift/oc v1.1.2-256-g2903c81 kubernetes v1.2.0-origin) The test result in step 3 is: the "Save" button is gray and prohibited, and there are prompts like "CPU request total for all containers is greater than pod maximum (500 millicores)", " CPU limit total for all containers is greater than pod maximum (500 millicores)" The bug is fixed, so move to VERIFIED.