Chapter 6. Bug fixes

Reconfiguration script does not install pull provider on push mode hosts after host content source change

Previously, when you changed the content source of a host and ran the manual reconfiguration script, the script attempted to install the foreman_ygg_worker package even if the host was in push mode for remote execution. This resulted in the failure of the script on hosts in push mode. Katello has been fixed to install the package on pull-mode hosts only. As a result, the reconfiguration script does not attempt to install the pull provider on hosts in push mode and the script succeeds.

hammer scap-content bulk-upload --type default now updates previously uploaded default compliance guidelines

Before this update, when an updated version of the default compliance guidelines was installed on the system where Satellite server was running, using the hammer scap-content bulk-upload --type default command did not update the guidelines on the Satellite server if the command was already used to upload a previous version of the guidelines before. As a consequence, Satellite in some cases did not include the latest SCAP content after hammer scap-content bulk-upload --type default was run. Now, when an updated version of the default compliance guidelines is available locally, running hammer scap-content bulk-upload --type default properly updates the guidelines on the Satellite server, even when the command was used to upload a previous version of the guidelines before.

Compute profile not applied on host update when using Hammer or API

Previously, when you used Hammer to change a host group associated with a different compute profile, Hammer attempted to update the hardware of your virtual machine. The Foreman core has been fixed to prevent compute profile application when a host is updated. As a result, when you change a host group of a virtual machine by using Hammer or API, Satellite no longer attempts to update the hardware. This behavior is now consistent with the Satellite web UI.

Provisioning uses correct hostname for RHEL 9 hosts

Previously, when you provisioned RHEL 9 to a host, Satellite registered the host with the localhost entry as the hostname. This resulted in consequent failures to provision other RHEL 9 hosts. The provisioning templates have been fixed to use the correct hostname. As a result, provisioning RHEL 9 hosts works as expected.

Specifying a custom organization label is now again possible in the Satellite web UI

Before this update, when creating a new organization in the Satellite web UI, the field to specify a custom label for the organization was missing. This update adds the missing field, and as a result, you can now specify a custom label when creating new organizations in the web UI.

RPM installation resolves pulpcore-api core dumps during concurrent repository synchronization

Previously, when synchronizing multiple large repositories to a Capsule concurrently, the pulpcore-api could crash and cause core dumps. Despite the crash, synchronization completed successfully. This issue is resolved by installing the python3.11-psycopg_c-3.2.3 RPM to avoid crashes.

Organization deletion failure caused by container push repositories

Previously, deleting an organization failed if any of its products included container push repositories. This issue occurred due to an error in the code that attempted to delete those repositories incorrectly which has been fixed. You can now delete organizations that include container push repositories.

On-demand repositories accessible after syncing

Previously, when you synchronized a repository by using the on-demand download policy and attempted to access the repository at the Published At URL, Satellite returned a 500 server error. Pulpcore has been fixed to handle content directory listing. As a result, you can access the repositories at the Published At URL .