Chapter 4. Known issues

Currently, the admin users are unable to deploy applications from a Dev Spaces workspace to OpenShift because those users' default projects are not the same as the Dev Spaces workspace project. This problem affects all devfiles that project an example command to deployment the application to OpenShift, including Node JS or Red Hat Fuse. This problem does not affect non-administrative user accounts on OpenShift because such users only have access to one project, which is the default <username>-devspaces namespace.

Currently, deploying the Vert.x Health Check Example application by running the 6-deploy-to-openshift command fails in disconnected environments. There is currently no workaround for this issue.

Currently, pulling of a private container image fails when the image is pulled from a container registry that was added in the dashboard, User Preferences Add Container Registry.

Currently, starting a workspace that uses the Technology Preview Visual Studio Code on OpenShift with a self-signed TLS certificate results in a blank browser tab.

Currently, a workspace might fail to start, and then you might be unable to delete it. This is caused by a persistent volume claim (PVC) issue. Subsequently, the same problem occurs with other workspaces.

Currently, using the Convert button in the OpenShift Dev Spaces dashboard to convert a Deprecated Node.js workspace with the node-debug or node-debug2 plug-in fails. The following error message is displayed:

Workspace conversion failed. Failed to create a new workspace from the devfile, reason: Unable to resolve theia plugins …​.

In OpenShift Dev Spaces 3.1, the node-debug and node-debug2 plug-ins have been updated to js-debug.

Currently, there is a known issue with the Java Gradle sample in restricted environments. Running the 1-build command to build an application might result in the FAILURE: Build failed with an exception error and failure to load a native library or not resolving a plug-in artifact in plug-in repositories. This issue has no workaround.

Currently, the Welcome To Your Workspace page in Che-Theia workspaces might appear blank when the workspace loads. This is caused by a missing self-signed TLS certificate in the browser.

Currently, the name of the CheCluster custom resource remains the same as before upgrading from CodeReady Workspaces 2.15 to OpenShift Dev Spaces 3.1. Customers whose Checluster custom resource name is codeready-workspaces before the upgrade will find the same name after the upgrade. This issue has no workaround.

When launching a sample application from the workspace, users can encounter an error message such as 502 Bad Gateway or application not available. The cause of this error is the Theia IDE displaying the notification that the application is ready before the application startup is complete.

Workaround: Wait a minute or two, and reload the browser tab.

See the related issue #21377.

In CodeReady Workspaces 2.15, users were able to edit their profile information in the Account page of the dashboard. Because OpenShift OAuth is now used exclusively for user management in OpenShift Dev Spaces 3.1, users can only edit their user profile within OpenShift. The Account page in the dashboard remains for display purposes only.

Currently, the per-workspace workspace storage strategy is not supported due to the change to the DevWorkspace engine. With migration from CodeReady Workspaces 2.15 to OpenShift Dev Spaces 3.1, existing workspaces change to the common strategy. Setting the PVC storage size other than 10 GB is currently not possible. This issue has no workaround.

When using the Cake-php sample, the Configure Apache Web Server DocumentRoots task fails with the following error: error sed: couldn’t open temporary file /etc/httpd/conf/sedSgv1Z4: Permission denied. There is currently no workaround for this issue.

Upgrading a CodeReady Workspaces 2.15 instance with the DevWorkspace engine enabled (Technology Preview) is simpler than the upgrade procedure described in the Administration Guide.

On IBM Power, the list of supported image streams is missing, which causes component creation to fail. There is currently no workaround for this issue.

The DEBUG panel displays No Configurations in the drop-down list because no configurations are loaded.

The OpenShift Connector plug-in fails to deploy because of the inability to access the odo image in the disconnected environment. There is no workaround for this issue.

When a task is running and there is some networking issue, the terminal window is cleared and contains no text. Even when the connection is restored, the terminal remains empty and loading. There is no workaround for this issue.

A workspace using Java with the JBoss EAP 7.3 devfile fails with the following error message: No delegateCommandHandler for vscode.java.startDebugSession. There is no workaround for this issue.

Currently, the following devfiles are not supported on IBM Z and IBM Power:

This issue has no workaround.

OpenShift Dev Spaces might fail to create a workspace when the network is unstable. OpenShift Dev Spaces displays an error such as Failed to run the workspace: "Waiting for pod 'workspace9fbid1gnx7273d47.maven-545f8c9cf4-hw79f' was interrupted." This issue has no workaround.

Currently, Golang-based workspaces do not include basic language server features such as code autocompletion.

On IBM Z and IBM Power, the debugging features cannot be activated in the Go workspace in OpenShift Dev Spaces 3.1. Delve, the required debugger for the Go programming language, is not available for these platforms. Because the debugger doesn’t exist, attempting to use it results in a Failed to continue error message. This issue has no workaround.