[ https://issues.jboss.org/browse/JBIDE-23136?page=com.atlassian.jira.plugi... ] Dmitrii Bocharov commented on JBIDE-23136: ------------------------------------------ I'm resolving this Jira as it is not a bug. h4. Explanation When we deploy image to Openshift with pushing it to the registry, first of all an Image stream is created. See the example of the part of image stream yaml: {code:yaml} dockerImageRepository: '172.30.1.1:5000/myproject/fedora' tags: - items: - created: '2017-11-08T11:44:23Z' dockerImageReference: >- 172.30.1.1:5000/myproject/fedora@sha256:d17b578fcd97f98a849d81783cd82674e229ebcb209c04b772dc5181dbc8a0a1 generation: 1 image: >- sha256:d17b578fcd97f98a849d81783cd82674e229ebcb209c04b772dc5181dbc8a0a1 tag: my-tag-2 {code} Although tag is specified, _dockerImageReference_ field will be used to pull image during the deployment. Look at the screenshot of the events: !pulling-by-digest.png|thumbnail! When you do pull by the digest, the tag is not specified. See docker documentation here: https://docs.docker.com/engine/reference/commandline/images/ (List image digests) and here https://docs.docker.com/engine/reference/commandline/pull/#pull-an-image-.... There is nothing wrong with it. Actually, it can be "fixed" by removing _removeTag_ in the _PushImageToRegistryJob_ class: {code:java} try { this.dockerConnection.removeTag(tmpImageName); } catch (DockerException | InterruptedException e) { return new Status(IStatus.WARNING, OpenShiftCoreActivator.PLUGIN_ID, "Pushed the selected Docker image into OpenShift registry but failed to untag it afterwards", e); } {code} This "untagging" is the equivalent to choosing the following 2 option in Eclipse Docker Tooling in push image wizard: !docker-tools-push-image-options.png|thumbnail! See [the comment|https://issues.jboss.org/browse/JBIDE-23136?focusedCommentId=1332...] of [~xcoulon] above for more details. However, i don't think it's necessary to remove the "untagging" because everything works correctly, there is nothing wrong with the <none> tag for image pulled by digest and because i agree to Xavier's comment, that we need to restore the initial state. So i'm resolving this jira by now. However if you have any more questions or have different opinion of how it must work, feel free to reopen and comment. -- This message was sent by Atlassian JIRA (v7.5.0#75005)