|Jive and High-Availability / Failover Behavior of HA Servers|
Any Jive deployment beyond a simple pilot or proof-of-concept must use file system storage, where each web application node reads and writes from a shared disk system via NFS.
In a single data center HA configuration, it's assumed that the external storage is redundant within the data center. For an illustration, see Designing a Single Data Center HA Configuration.
In the event of a local (within the single data center) or a catastrophic (the entire data center) failure, it is assumed that the storage layer that Jive is configured with is redundant and that recovery is handled transparently by the underlying storage system.
In a multiple data center HA configuration, it's assumed that the external storage is being replicated across the data centers transparently. For an illustration, see Designing a Multiple Data Center HA Configuration.