Provide customers more stable, flexible, reliable and faster storage service deployments with more reliable and flexible lifecycle management and possibility to rollback when something goes wrong.
Moving back to POST; there are a few important parts missing/not working correctly yet. 1. SwiftRawDisks Not supported yet; the defined disks won't be properly setup & mounted. Additionally this enables the mount_check parameter, which is not working inside a containerized environment the way it is used. 2. Containerized mounts Swift running inside the container is not able to detect wether a device is mounted or not. While mapping the devices into the container allows correct mount checks, it is not an option because multiple containers need to access the same disk (for example object-server and object-replicator). 3. Ring management Storing and retrieving the overcloud rings from the undercloud does no longer work. This is essential to keep the rings consistent across the cluster even if nodes are replaced or added later. There might be a few other things as well; however I'm working on the above topics right now.
Proposed patch for task #1: https://review.openstack.org/#/c/466419/
All required upstream patches merged -> POST.
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, 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/RHEA-2017:3462