Cluster redundancy

The redundancy ensure that the cluster resources (VM, containers , etc) are always accessible and operational on the cluster. Various clusters configurations are possible, among which:

Active/active cluster: the resources are duplicated distributed along cluster nodes.
Active/passive cluster: the resources are distributed on a single node, redounded by a recovery machine in case of a failure.

To get a better view of these configuration lets take a look applied to SEAPATH:

Resource strategy

Resource strategy
Active/Active (many instances)	Active/Passive (single instance)
Many VMs instances executed in the same time Higher hardware resources usage Implementation example : Two VMs executed in parallel, without synchronization Lower service downtime Only resource downtime detection, no resource restart needed	A single VM instance executed in the same time Minimization of the needed hardware resources Higher service downtime during a migration VM downtime detection VM restart on another cluster node