configurations
The HA-Ops tools must operate consistently on different configurations, and should support migration between configurations.
An Embedded GraphDB deployment is a Client application using Neo4j as a library, running on a Java Virtual Machine.
An Embedded deployment can manually migrate to a Server or scale to a High-Availability configuration.
serve E = S
scale E = HA
The Server deployment separates the Client runtime from the GraphDB, which provides access through a Proxy (REST, wire protocol, etc).
A Server can be scaled to a High-Availability Cluster (HAC).
scale S = HAC
Server and simplified view
An Embedded High-Availability provides read scaling that must be managed entirely by the client. (No tool support)
Embedded HA can manually migrate to a High-Availability Server Cluster (HAC).
serve HA = HAC
A High Availability Cluster provides read scaling for a Server configuration. The client manages connections to each of the HAC member instances.
HAC currently has no forward migrations, though a Managed Proxy would be the next logical addition.
High-Availability Cluster and simplified view