Masakari: provides Instances High Availability Service for OpenStack
recover KVM-based Virtual Machine(VM)s from failure events
an API service to manage and control the automated rescue mechanism
Cluster stack: Corosync, Pacemaker, Pacemaker-remote for HA
Corosync: cluster engine
Pacemaker: resource manager
Pacemaker-remote: lightweight resource manager for scalability.
(corosync/pacemaker cluster has a limit of 16 nodes.)
Controller nodes
corosync/pacemaker
masakari-api: Get notifications of instance/host failures from monitor
masakari-engine: Process recovery workflow for instance/host failure
Compute nodes
pacemaker-remote
masakari-instancemonitor
masakari-hostmonitor
masakari-processmonitor is not used by OkiDoki
because openstack components are running as pods on kubernetes system.
Kubernetes itself can handle nova-compute service failure.