metadata:
  annotations:
    prow.k8s.io/context: ci/prow/e2e-openshift-aws-serial
    prow.k8s.io/job: pull-scylla-operator-master-e2e-openshift-aws-serial
  creationTimestamp: "2025-10-07T11:08:43Z"
  generation: 10
  labels:
    created-by-prow: "true"
    event-GUID: fb42b0e0-a36d-11f0-8228-9e55713c9ea2
    prow.k8s.io/build-id: "1975518647317172224"
    prow.k8s.io/context: e2e-openshift-aws-serial
    prow.k8s.io/id: 4554cb53-8681-48ab-8e8c-ab7354e0084b
    prow.k8s.io/is-optional: "true"
    prow.k8s.io/job: pull-scylla-operator-master-e2e-openshift-aws-serial
    prow.k8s.io/refs.base_ref: master
    prow.k8s.io/refs.org: scylladb
    prow.k8s.io/refs.pull: "2987"
    prow.k8s.io/refs.repo: scylla-operator
    prow.k8s.io/type: presubmit
  name: 4554cb53-8681-48ab-8e8c-ab7354e0084b
  namespace: prow-workspace
  resourceVersion: "1759842739211423018"
  uid: 4257c792-4104-428d-b736-a14e294a8558
spec:
  agent: kubernetes
  cluster: default
  context: ci/prow/e2e-openshift-aws-serial
  decoration_config:
    gcs_configuration:
      bucket: gs://scylla-operator-prow
      path_strategy: explicit
    gcs_credentials_secret: gcs-credentials
    github_api_endpoints:
    - http://ghproxy.prow.svc
    - https://api.github.com
    github_app_id: "112385"
    github_app_private_key_secret:
      key: cert
      name: github-token
    grace_period: 15m0s
    resources:
      clonerefs:
        requests:
          cpu: 100m
      initupload:
        requests:
          cpu: 100m
      place_entrypoint:
        requests:
          cpu: 100m
      sidecar:
        requests:
          cpu: 100m
    timeout: 4h0m0s
    utility_images:
      clonerefs: us-docker.pkg.dev/k8s-infra-prow/images/clonerefs:v20250522-9b3f5faca
      entrypoint: us-docker.pkg.dev/k8s-infra-prow/images/entrypoint:v20250522-9b3f5faca
      initupload: us-docker.pkg.dev/k8s-infra-prow/images/initupload:v20250522-9b3f5faca
      sidecar: us-docker.pkg.dev/k8s-infra-prow/images/sidecar:v20250522-9b3f5faca
  job: pull-scylla-operator-master-e2e-openshift-aws-serial
  namespace: prow-workspace
  pod_spec:
    containers:
    - args:
      - |
        source "./hack/lib/kube.sh"

        function cleanup {
          touch /tmp/shared/setup.finished
          kubectl -n ci-clusters delete --wait=false --ignore-not-found kubernetescluster/"${POD_UID}"
        }
        trap cleanup EXIT

        pull_refs_hash=$( sha512sum <<< "${PULL_REFS}" | cut -c 1-31 )
        SOCI_PR_TAG="pr-${pull_refs_hash}"
        export SOCI_PR_TAG

        # Don't create e2e cluster before an image is build, so we don't waste infra when it doesn't compile.
        ERR_LOG=$( mktemp )
        export ERR_LOG
        timeout 15m bash -xc 'until skopeo inspect --format='"'"'{{ printf "%s@%s" .Name .Digest }}'"'"'  docker://"quay.io/scylladb-dev/ci:scylla-operator-${SOCI_PR_TAG}" 2>>"${ERR_LOG}"; do sleep 5; done' || ( cat "${ERR_LOG}" && exit 2 )

        kubectl version -o yaml

        kubectl -n ci-clusters create -f - <<EOF
        apiVersion: ci.scylladb.com/v1alpha1
        kind: KubernetesCluster
        metadata:
          name: "${POD_UID}"
        spec:
          version: quay.io/okd/scos-release:4.19.0-okd-scos.9
          ttlSeconds: 14400
          credentialsSecret:
            name: "${POD_UID}-kubeconfig"
          type: OpenShift
          openshift:
            platformType: AWS
            credentialsSecret:
              name: aws-openshift-admin-creds
            installConfig:
              baseDomain: okd.ci.aws.scylla-operator.scylladb.com
              platform:
                aws:
                  region: us-east-1
                  userTags:
                    keep: "8"
              controlPlane:
                name: master
                platform:
                  aws:
                    zones:
                    - us-east-1a
                    type: m6i.xlarge
                replicas: 3
              compute:
              - name: worker
                platform:
                  aws:
                    zones:
                    - us-east-1a
                    type: m5a.2xlarge
                replicas: 2
        EOF

        wait-for-object-creation ci-clusters configmap/"${POD_UID}-openshift-install-config"

        mkdir -p "${ARTIFACTS}"/.ci/manifests/cluster
        kubectl -n ci-clusters get configmap/"${POD_UID}-openshift-install-config" -o=yaml > "${ARTIFACTS}/.ci/manifests/cluster/${POD_UID}-openshift-install-config.cm.yaml"

        timeout -v 2h bash -c 'until kubectl -n ci-clusters wait --for=condition=Degraded=False kubernetescluster/"${POD_UID}" --timeout=2h && kubectl -n ci-clusters wait --for=condition=Progressing=False kubernetescluster/"${POD_UID}" --timeout=2h && kubectl -n ci-clusters wait --for=condition=ClusterBootstrapped=True kubernetescluster/"${POD_UID}" --timeout=2h; do sleep 1; done'

        kubectl -n ci-clusters get secret/"${POD_UID}-kubeconfig" --template='{{ .data.kubeconfig }}' | base64 -d > /tmp/shared/kubeconfig.init
        kubectl --kubeconfig=/tmp/shared/kubeconfig.init config set-context --current --namespace 'default-unexisting-namespace'

        # Sanity check.
        kubectl --kubeconfig=/tmp/shared/kubeconfig.init version -o yaml
        kubectl --kubeconfig=/tmp/shared/kubeconfig.init config view

        kubectl --kubeconfig=/tmp/shared/kubeconfig.init label machineconfigpool worker cpumanager-policy=static
        kubectl --kubeconfig=/tmp/shared/kubeconfig.init create -f - <<EOF
        apiVersion: machineconfiguration.openshift.io/v1
        kind: KubeletConfig
        metadata:
          name: cpumanager-policy-static
        spec:
          machineConfigPoolSelector:
            matchLabels:
              cpumanager-policy: static
          kubeletConfig:
            cpuManagerPolicy: static
        EOF

        # Wait until KubeletConfig is reconciled and applied on matching machine config pools.
        kubectl --kubeconfig=/tmp/shared/kubeconfig.init wait --for=condition=Updating=True machineconfigpool/worker --timeout=10m
        kubectl --kubeconfig=/tmp/shared/kubeconfig.init wait --for=condition=Updated=True machineconfigpool/worker --timeout=30m

        # Signal cluster bootstrap by sharing the kubeconfig.
        mv /tmp/shared/kubeconfig{.init,}

        set +x
        # TODO: Wait on active signal like updating a file every X seconds
        #       so we can deal with the other container being OOM killed.
        echo "Waiting for test to finish..."
        until [[ -f "/tmp/shared/test.finished" ]]; do sleep 1; done
        echo "Test has finished."
        set -x
      command:
      - /usr/bin/bash
      - -euExo
      - pipefail
      - -O
      - inherit_errexit
      - -c
      env:
      - name: POD_UID
        valueFrom:
          fieldRef:
            fieldPath: metadata.uid
      image: quay.io/scylladb/scylla-operator-images:kube-tools
      imagePullPolicy: Always
      name: setup
      resources:
        limits:
          cpu: 100m
          memory: 200Mi
      securityContext:
        allowPrivilegeEscalation: false
        capabilities:
          drop:
          - ALL
        seccompProfile:
          type: RuntimeDefault
      volumeMounts:
      - mountPath: /var/run/secrets/kubernetes.io/serviceaccount/
        name: kube-api-access
      - mountPath: /tmp/shared
        name: shared-data
    - args:
      - |
        trap 'touch /tmp/shared/test.finished' EXIT

        set +x
        echo "Waiting for cluster to be provisioned..."
        until [[ -f "${KUBECONFIG}" || -f /tmp/shared/setup.finished ]]; do sleep 1; done
        if [[ -f /tmp/shared/setup.finished ]]; then
          echo "Cluster provisioning failed. Exiting."
          exit 1
        fi
        echo "Cluster provisioning has finished."
        set -x

        # TODO(rzetelskik): move labelling to KubernetesCluster manifest once we support dedicated node pools
        # Xref: https://github.com/scylladb/scylla-operator-release/issues/287
        kubectl label node --selector=node-role.kubernetes.io/worker="" scylla.scylladb.com/node-type=scylla

        # Setup info and kubeconfig sanity check
        kubectl config view
        kubectl version -o yaml
        kubectl cluster-info

        pull_refs_hash=$( sha512sum <<< "${PULL_REFS}" | cut -c 1-31 )
        SOCI_PR_TAG="pr-${pull_refs_hash}"

        SO_SUITE=scylla-operator/conformance/serial
        export SO_SUITE
        SO_NODECONFIG_PATH=""
        export SO_NODECONFIG_PATH
        SO_CSI_DRIVER_PATH=""
        export SO_CSI_DRIVER_PATH
        SO_IMAGE=quay.io/scylladb-dev/ci:scylla-operator-${SOCI_PR_TAG}
        export SO_IMAGE
        SO_SCYLLACLUSTER_NODE_SERVICE_TYPE=Headless
        export SO_SCYLLACLUSTER_NODE_SERVICE_TYPE
        SO_SCYLLACLUSTER_NODES_BROADCAST_ADDRESS_TYPE=PodIP
        export SO_SCYLLACLUSTER_NODES_BROADCAST_ADDRESS_TYPE
        SO_SCYLLACLUSTER_CLIENTS_BROADCAST_ADDRESS_TYPE=PodIP
        export SO_SCYLLACLUSTER_CLIENTS_BROADCAST_ADDRESS_TYPE
        SO_SCYLLACLUSTER_STORAGECLASS_NAME=""
        export SO_SCYLLACLUSTER_STORAGECLASS_NAME
        SO_E2E_TIMEOUT=60m
        export SO_E2E_TIMEOUT
        timeout --verbose --signal INT --kill-after=100m 90m ./hack/.ci/run-e2e-openshift-aws.sh
      command:
      - /usr/bin/bash
      - -euExo
      - pipefail
      - -O
      - inherit_errexit
      - -c
      env:
      - name: POD_UID
        valueFrom:
          fieldRef:
            fieldPath: metadata.uid
      - name: KUBECONFIG
        value: /tmp/shared/kubeconfig
      image: quay.io/scylladb/scylla-operator-images:kube-tools
      imagePullPolicy: Always
      name: test
      resources:
        limits:
          cpu: 100m
          memory: 1200Mi
      securityContext:
        allowPrivilegeEscalation: false
        capabilities:
          drop:
          - ALL
        seccompProfile:
          type: RuntimeDefault
      volumeMounts:
      - mountPath: /tmp/shared
        name: shared-data
    serviceAccountName: e2e
    volumes:
    - emptyDir: {}
      name: shared-data
    - name: kube-api-access
      projected:
        defaultMode: 400
        sources:
        - serviceAccountToken:
            path: token
        - configMap:
            items:
            - key: ca.crt
              path: ca.crt
            name: kube-root-ca.crt
        - downwardAPI:
            items:
            - fieldRef:
                apiVersion: v1
                fieldPath: metadata.namespace
              path: namespace
  prowjob_defaults:
    tenant_id: GlobalDefaultID
  refs:
    base_link: https://github.com/scylladb/scylla-operator/commit/1fe085b2229c4faaf511fa8705a8c34d9a093094
    base_ref: master
    base_sha: 1fe085b2229c4faaf511fa8705a8c34d9a093094
    org: scylladb
    path_alias: github.com/scylladb/scylla-operator
    pulls:
    - author: czeslavo
      author_link: https://github.com/czeslavo
      commit_link: https://github.com/scylladb/scylla-operator/pull/2987/commits/cb292ece799f1339d112d720690753259fb120bc
      head_ref: use-in-container-device-path
      link: https://github.com/scylladb/scylla-operator/pull/2987
      number: 2987
      sha: cb292ece799f1339d112d720690753259fb120bc
      title: '[WIP] Fix remaining loop devices manipulations in NodeSetup E2Es'
    repo: scylla-operator
    repo_link: https://github.com/scylladb/scylla-operator
  report: true
  rerun_command: /test e2e-openshift-aws-serial
  type: presubmit
status:
  build_id: "1975518647317172224"
  completionTime: "2025-10-07T13:12:16Z"
  description: Job failed.
  pendingTime: "2025-10-07T11:08:43Z"
  pod_name: 4554cb53-8681-48ab-8e8c-ab7354e0084b
  prev_report_states:
    gcsk8sreporter: failure
    gcsreporter: failure
    github-reporter: failure
  startTime: "2025-10-07T11:08:43Z"
  state: failure
  url: https://prow.scylla-operator.scylladb.com/view/gs/scylla-operator-prow/pr-logs/pull/scylladb_scylla-operator/2987/pull-scylla-operator-master-e2e-openshift-aws-serial/1975518647317172224