We need to update and expand the information that begins here:
http://docs.mirantis.com/openstack/fuel/master/reference-architecture.html#details-of-multi-node-with-ha-deployment
Some topics to cover:
- Describe single-controller HA mode
- Internal details of how a single-controller environment can be converted to a three-or-more Controller environment (true HA)
- Technical details (including File Format reference Pages) for howFuel deploys and redeploys an HA environment.
Include information about viewing Puppet manifests, etc.
- It would be really nice to just list the order in which components are deployed, perhaps showing how to follow the progress
in system logs
- Graphics need to always show 1, 3, 5... Controllers, not 2 Controllers as is currently depicted
- Details of failover for all components -- Controllers, Compute Nodes, Ceph-OSD clusters, MongoDB clusters, etc
Evgeny is working on HA internal documentation.
What's more, Tomasz wanted to discuss HA issues...I should ping him so that we could plan our activities.