Skip to content

Conversation

@tuxerrante
Copy link
Collaborator

Which issue this PR addresses:

Fixes failures to install test clusters without hive when running a containerized development RP.

Based on #4509

What this PR does / why we need it:

When using a containerized RP to deploy a cluster, and running a non-hive deployment (local openshift installer container), the RP fails to connect to the hosts podman socket to create the openshift installer.
This PR makes sure the RP container starts with the necessary mounts and permissions to create said installer.

Test plan for issue:

Deploy a cluster against the containerized RP with hive env variables unset.

Is there any documentation that needs to be updated for this PR?

This PR also updates the necessary documentation to make sure users are running podman compose.

How do you know this will function as expected in production?

Not intended for use in production.

@tuxerrante tuxerrante assigned tuxerrante and aasserzo and unassigned tuxerrante Dec 12, 2025
…RET_SA_ACCOUNT_NAME from the env file if not set. This improves robustness in the secrets and secrets-update targets.
…volume path for podman socket to use a default value if XDG_RUNTIME_DIR is not set, and included FEDORA_REGISTRY as a build argument for better flexibility.
…gument for improved flexibility. Removed outdated image update automation from Makefile to streamline the build process.
it was not possible to push an arm64 image to the intsvc CR
On MacOS and docker userns doesn't work as expected. It is more safe to
explicitly pass the UID if needed. By symlinking the socket path we
could probably use only one path mount here
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants