Description of problem: On Openshift 4.3.0 I cannot create application as Developer from Catalog when I use image URLs from Docker hub. Steps to Reproduce: As developer click on +Add. Then, select 'Container Image'. Insert an URL from Docker hub, in my case 'tensorflow/tensorflow:latest-jupyter'. Now create a deployment. The pods cannot be started because the image tag is incorrectly resolved to 'tensorflow:'. [root@worker01 ~]# oc get pods -n przemek NAME READY STATUS RESTARTS AGE tensorflow-cb48f6c67-c8vp5 0/1 InvalidImageName 0 32s [root@worker01 ~]# oc logs tensorflow-cb48f6c67-c8vp5 -n przemek Error from server (BadRequest): container "tensorflow" in pod "tensorflow-cb48f6c67-c8vp5" is waiting to start: InvalidImageName [root@worker01 ~]# oc describe pod tensorflow-cb48f6c67-c8vp5 -n przemek Name: tensorflow-cb48f6c67-c8vp5 Namespace: przemek Priority: 0 Node: worker02.ocp4.scale.com/192.168.1.16 Start Time: Tue, 11 Feb 2020 15:49:28 +0100 Labels: app=tensorflow deploymentconfig=tensorflow pod-template-hash=cb48f6c67 Annotations: k8s.v1.cni.cncf.io/networks-status: [{ "name": "openshift-sdn", "interface": "eth0", "ips": [ "10.128.2.58" ], "dns": {}, "default-route": [ "10.128.2.1" ] }] openshift.io/generated-by: OpenShiftWebConsole openshift.io/scc: restricted Status: Pending IP: 10.128.2.58 IPs: IP: 10.128.2.58 Controlled By: ReplicaSet/tensorflow-cb48f6c67 Containers: tensorflow: Container ID: Image: tensorflow: Image ID: Port: 8888/TCP Host Port: 0/TCP State: Waiting Reason: InvalidImageName Ready: False Restart Count: 0 Environment: <none> Mounts: /var/run/secrets/kubernetes.io/serviceaccount from default-token-nz6m4 (ro) Conditions: Type Status Initialized True Ready False ContainersReady False PodScheduled True Volumes: default-token-nz6m4: Type: Secret (a volume populated by a Secret) SecretName: default-token-nz6m4 Optional: false QoS Class: BestEffort Node-Selectors: <none> Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s node.kubernetes.io/unreachable:NoExecute for 300s Events: Type Reason Age From Message ---- ------ ---- ---- ------- Normal Scheduled <unknown> default-scheduler Successfully assigned przemek/tensorflow-cb48f6c67-c8vp5 to worker02.ocp4.scale.com Warning InspectFailed 8s (x8 over 79s) kubelet, worker02.ocp4.scale.com Failed to apply default image tag "tensorflow:": couldn't parse image reference "tensorflow:": invalid reference format Warning Failed 8s (x8 over 79s) kubelet, worker02.ocp4.scale.com Error: InvalidImageName
Using a URL like `tensorflow/tensorflow:latest-jupyter` should work as is. However as a work-around you can try `docker.io/tensorflow/tensorflow:latest-jupyter`.
No luck with the URL you provided. I got the same error messages. [root@worker01 ~]# oc get pods -n przemek NAME READY STATUS RESTARTS AGE tensorflow-cb48f6c67-dgmqp 0/1 InvalidImageName 0 54s [root@worker01 ~]# oc logs tensorflow-cb48f6c67-dgmqp -n przemek Error from server (BadRequest): container "tensorflow" in pod "tensorflow-cb48f6c67-dgmqp" is waiting to start: InvalidImageName [root@worker01 ~]# oc describe pod tensorflow-cb48f6c67-dgmqp -n przemek Name: tensorflow-cb48f6c67-dgmqp Namespace: przemek Priority: 0 Node: worker02.ocp4.scale.com/192.168.1.16 Start Time: Tue, 11 Feb 2020 17:21:06 +0100 Labels: app=tensorflow deploymentconfig=tensorflow pod-template-hash=cb48f6c67 Annotations: k8s.v1.cni.cncf.io/networks-status: [{ "name": "openshift-sdn", "interface": "eth0", "ips": [ "10.128.2.69" ], "dns": {}, "default-route": [ "10.128.2.1" ] }] openshift.io/generated-by: OpenShiftWebConsole openshift.io/scc: restricted Status: Pending IP: 10.128.2.69 IPs: IP: 10.128.2.69 Controlled By: ReplicaSet/tensorflow-cb48f6c67 Containers: tensorflow: Container ID: Image: tensorflow: Image ID: Port: 8888/TCP Host Port: 0/TCP State: Waiting Reason: InvalidImageName Ready: False Restart Count: 0 Environment: <none> Mounts: /var/run/secrets/kubernetes.io/serviceaccount from default-token-nz6m4 (ro) Conditions: Type Status Initialized True Ready False ContainersReady False PodScheduled True Volumes: default-token-nz6m4: Type: Secret (a volume populated by a Secret) SecretName: default-token-nz6m4 Optional: false QoS Class: BestEffort Node-Selectors: <none> Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s node.kubernetes.io/unreachable:NoExecute for 300s Events: Type Reason Age From Message ---- ------ ---- ---- ------- Normal Scheduled <unknown> default-scheduler Successfully assigned przemek/tensorflow-cb48f6c67-dgmqp to worker02.ocp4.scale.com Warning InspectFailed 14s (x7 over 69s) kubelet, worker02.ocp4.scale.com Failed to apply default image tag "tensorflow:": couldn't parse image reference "tensorflow:": invalid reference format Warning Failed 14s (x7 over 69s) kubelet, worker02.ocp4.scale.com Error: InvalidImageName
It's interesting to see the deploymentconfig, what does `oc get dc tensorflow -o yaml` print?
Tested in 4.4 as well and observed the same failure. Tag is incorrectly set when choosing to create Deployments. It works when creating a Deployment Config. Fails again when creating a Knative service: `"image-registry.openshift-image-registry.svc:5000/cvogt/tensorflow-kn:latest-jupyter": failed to resolve image to digest: failed to fetch image information: MANIFEST_UNKNOWN: manifest unknown; map[Tag:latest-jupyter].`
Why is this marked as a test blocker?
I would like to test specific workloads with docker images. These are blocked as for now with deployments.
Verified on ci build: 4.4.0-0.ci-2020-02-18-102514 Chrome browser: 76.0.3809.132
Created attachment 1663900 [details] Using Docker images for Applications is possible in Openshift 4.4.0
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-2020:0581
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days