Uploaded image for project: 'Red Hat OpenShift Data Science'
  1. Red Hat OpenShift Data Science
  2. RHODS-5761

Failed to fetch GPU: HTTP request failed

XMLWordPrintable

    • False
    • None
    • False
    • Testable
    • Yes
    • 1.19.0-14
    • No
    • No
    • Yes
    • None
    • RHODS 1.19, RHODS 1.20

      When installing RHODS 1.19.0-7, this error message appears on the dashboard notebook spawner page:

      cause by this underlying HTTP error when trying to get the /api/gpu endpoint

      Workarounds

      The issue may be resolved by adding a new ClusterRole:

      kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata: 
  name: rhods-dashboard-autoscaling
rules: 
  - verbs: 
      - get
      - list
    apiGroups: 
      - machine.openshift.io
      - autoscaling.openshift.io
    resources: 
      - machineautoscalers
      - machinesets

      And a new ClusterRoleBinding:

      kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: rhods-dashboard-autoscaling
subjects:
  - kind: ServiceAccount
    name: rhods-dashboard
    namespace: redhat-ods-applications
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: rhods-dashboard-autoscaling

       

       

        1. image-2022-11-04-13-53-13-897.png
          image-2022-11-04-13-53-13-897.png
          78 kB
        2. image-2022-11-04-13-53-44-306.png
          image-2022-11-04-13-53-44-306.png
          15 kB
        3. Screenshot from 2022-11-04 10-30-30.png
          Screenshot from 2022-11-04 10-30-30.png
          158 kB
        4. image-2022-11-11-14-26-36-618.png
          image-2022-11-11-14-26-36-618.png
          198 kB
        5. rhods-dashboard-c79597765-brvhh-rhods-dashboard.log
          767 kB

              rh-ee-mroman Maros Roman (Inactive)
              kpouget2 Kevin Pouget
              Votes:
              0 Vote for this issue
              Watchers:
              9 Start watching this issue

                Created:
                Updated:
                Resolved: