Description of problem: On bigger environment, connections to the OVN databases may take a lot of time. Neutron OVN mechanism attempts to connect in the pre-fork to configure OVN DB server but API workers are started before. API workers rely on hash ring mechanism that considers hash ring node dead if it's not been touched in the last 60 seconds. The touching mechanism is done by periodic task that's executed from each maintenance worker, however the worker is started in a post-fork. This means, if connection to both OVN DBs take more than 60 seconds, the API worker hash ring nodes are considered dead. It eventually recovers but it takes more than an hour. This is a regression caused by https://review.opendev.org/c/openstack/networking-ovn/+/812304 Neutron shouldn't be configuring other services. Version-Release number of selected component (if applicable): 16.1.8 How reproducible: Always Steps to Reproduce: 1. Start neutron server 2. 3. Actual results: Expected results: Additional info:
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Red Hat OpenStack Platform 16.1.9 bug fix and enhancement advisory), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2022:8795