Description of problem: With the introduction of the change as instructed in [1] point number 8, under /etc/hosts RHOSP16 undercloud installation fails at step 5. Rabbitmq will try to setup cluster on the first IP address hostname mapping as mentioned in /etc/hosts file and the installation fails as rabbitmq would fail to run on public host IP and ansible block will append the hostname entries. This change is important as any user who modifies /etc/hosts file will encounter a failure in undercloud installation. Additional Note: The statement might itself be confusing and the end user may consider it as an entry for public routed network. So we should specify (undercloud_public_host) while writing this statement. Resolution: 1. Skip or comment this entry in /etc/hosts before running undercloud install. 2. Have an alternative plan to introduce this IP address hostname mapping (for public host) via ansible block so that it is appended. [1] - https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/16.0/html/director_installation_and_usage/preparing-for-director-installation#preparing-the-undercloud Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: Introduce public host IP and hostname mapping as mentioned in [1] point number 8. Run undercloud installation again. Actual results: Undercloud installation fails at Step 5. Expected results: Undercloud installation succeeds successfully. Additional info:
Also, I missed adding in description earlier that after commenting the entry in /etc/hosts we will need to perform "podman restart rabbitmq" , before running undercloud install again.
The workaround is mentioned in the first comment. We are working on a proper fix in Director via RHBZ #1792422. *** This bug has been marked as a duplicate of bug 1792422 ***