Description of problem: Setting Build Defaults (requests/limits) has no effect on actual build pod. Version-Release number of selected component (if applicable): 4.1.4 How reproducible: Always Steps to Reproduce: 1. Set up Build Defaults oc describe build.config.openshift.io/cluster Name: cluster Namespace: Labels: <none> Annotations: release.openshift.io/create-only: true API Version: config.openshift.io/v1 Kind: Build Metadata: Creation Timestamp: 2019-07-14T14:51:04Z Generation: 4 Resource Version: 570536 Self Link: /apis/config.openshift.io/v1/builds/cluster UID: cdbbd18d-a646-11e9-a64f-0206c3a2392c Spec: Build Defaults: Resources: Limits: Cpu: 1 Memory: 3Gi Requests: Cpu: 1 Memory: 2Gi Events: <none> 2. Set up a LimitRange (in a Project Request Template) and create a new project oc describe limits wk-builds-core-resource-limits Name: wk-builds-core-resource-limits Namespace: wk-builds Type Resource Min Max Default Request Default Limit Max Limit/Request Ratio ---- -------- --- --- --------------- ------------- ----------------------- Container memory 6Mi 6Gi 512Mi 1Gi - Container cpu - - 250m 500m - Pod memory 6Mi 12Gi - - - 2. Start a new build oc new-app https://github.com/wkulhanek/rhte-app.git 3. Check the build pod oc describe pod rhte-app-1-build [...] Limits: cpu: 500m memory: 1Gi Requests: cpu: 250m memory: 512Mi [...] Actual results: The pod uses the defaults from the Limit Range rather than the build defaults. Expected results: The pod uses the defaults from the build defaults therefore not needing to fall back on the defaults from the limit range
Marking this as a regression from 3.11, blocker for 4.2.0
FYI, @Wolfgang Kulhanek your Build Default config incorrect, should be Spec: buildDefaults: Resources:
@wewang My build defaults are straight from the docs: https://docs.openshift.com/container-platform/4.1/builds/build-configuration.html This is exactly what I have: apiVersion: config.openshift.io/v1 kind: Build metadata: annotations: release.openshift.io/create-only: "true" creationTimestamp: "2019-07-14T14:51:04Z" generation: 4 name: cluster resourceVersion: "570536" selfLink: /apis/config.openshift.io/v1/builds/cluster uid: cdbbd18d-a646-11e9-a64f-0206c3a2392c spec: buildDefaults: resources: limits: cpu: 1 memory: 3Gi requests: cpu: 1 memory: 2Gi I don't see a difference to what you are writing. The only difference I see is "Resources" in your example vs "resources" in mine and the docs.
Oh I see. I submitted the "oc describe" and you gave me YAML. So the YAML should be fine.
@Ricardo: I keep getting notification that info is needed from me - but I don't see what. Do you actually need something from me?
@Ricardo Maraschini After waiting ds rollout, config applied to build pod, thanks version: 4.2.0-0.nightly-2019-08-07-214151 $oc describe pod/ruby-hello-world-1-build Limits: cpu: 500m memory: 1Gi Requests: cpu: 250m memory: 1Gi
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:2922