In the gateway architecture today, gateways register with a primary call controller; however, the IP interface through which the gateway registers through a procr interface in the case where the call controller is a Duplicated server series or supported servers (through the Processor Ethernet interface). For supported servers, see Avaya Aura® Communication Manager Hardware Description and Reference.

The Survivable Core Server feature provides survivability to Media gateways and Media servers. Using the Survivable Core Server feature, backup servers can be placed in various locations in the customer’s network. The backup servers supply service to Port Networks in the case where the supported server, or the Duplicated series server pair fails, or connectivity to the main Communication Manager server is lost. Survivable Core Servers can be either supported servers or Duplicated series servers, and offer full Communication Manager functionality when in survivable mode, provided sufficient connectivity exists to other Avaya components (for example, endpoints, gateways, and messaging servers).

When designing a network to support Survivable Core Servers, consider the following:

• Survivable Core Servers register to the main server(s) through a procr. Each Survivable Core Servers must be able to communicate with a procr to download translations.

The gateway cannot distinguish between registration through a procr or registration to a server directly. Prior to Communication Manager 3.0, without Survivable Core Servers, if a gateway successfully registered with a primary call controller IP address, then the gateway was properly registered with the primary call controller. However, in Communication Manager 3.0 and later, when a gateway completes a successful registration through an IP address defined as a primary call controller address, if that address is a procr, the gateway may not necessarily be registered with the true primary call controller. The port network that houses the procr may be under control of a Survivable Core Servers; but the gateway will not know that it is registered with a Survivable Core Servers.

When the traditional port network migrates back to the primary call controller, then the gateway loses its H.248 link, and the Link Loss Recovery algorithm engages, and that should be sufficient. The Auto Fallback to Primary feature only engages if the gateway drops the connection and registers with a Survivable Remote Servers. The Survivable Core Servers migration should only occur if the port network is reasonably certain to return to the primary call controller, so the gateway would simply return to the same procr interface. Now, when the gateway returns to the same procr interface, the Link Loss Recovery feature performs a context audit with the primary controller and learns that the primary call controller is not aware of the gateway. The controller in this case issues a warm start request to the gateway, or potentially different behavior if connection preservation is active at the same time. The Auto-Fallback feature is not affected by Survivable Core Servers.

For more information on survivable core servers, see Avaya Aura® Communication Manager Survivable Options.