Terminology

A data center and the Transaction Engine (TE) and Storage Manager (SM) processes in it where the database workload runs. A database requires at least one active data center, containing at least one TE and at least one SM.

All database changes are replicated to a remote passive data center without the database slowing down because of inter-datacenter delay in communication between remote locations. In the event of a disaster, a recent, consistent snapshot of the database contents can be recovered and quickly put into service.

A Storage Manager (SM) that is a passive observer of all storage groups that it serves. It passively receives changes to a storage group but does not participate in NuoDB’s performance-sensitive network protocols. Transactions in the active data center do not depend on the latency for round-trip communication between the TEs and ASMs. So, latency between ASMs and other nodes does not affect the database performance.

The number of bytes per second that a network connection can transmit.

The delay from when a message is sent into a network connection until the other end receives it.

The procedure for passing control of a database to a passive data center. It becomes the new active data center. Handoff is the response to a disaster that has eliminated all active data centers. This resets the database to the most recent consistent state available in archives in the passive data center.

See Asynchronous Storage Manager (ASM).

An additional data center containing Asynchronous Storage Manager processes that passively receive and record database changes. You can set up one or more passive data centers in remote locations for disaster recovery.

Transmission of database changes over the network from one TE or SM process to others.