High Availability and Geo-Redundancy

Last Updated : Dec 14, 2020 |
Prolog information
Metadata
Keywords:
An IP network exists between the following components:
  • UNIStim endpoints or MGC of the TDM endpoints and the Device Adapter nodes (Avaya Breeze® platform nodes where the Device Adapter Snap-In is installed):
    • An IP network exists between the UNIStim endpoints and the Device Adapter nodes in the cluster.
      An IP connection failure may occur between a UNIStim endpoint and the Device Adapter node serving the endpoint.
    • An IP network exists between the MGC of the TDM endpoints (digital and analog endpoints) and the Device Adapter nodes in the cluster.
      An IP connection failure may occur between an MGC serving multiple TDM endpoints and the Device Adapter node serving the MGC.
  • An IP network exists between the Avaya Breeze® platform cluster with the Device Adapter nodes and the remainder of the Avaya Aura® components.
    An IP connection failure may occur between the Avaya Breeze® platform cluster and the Avaya Aura® components.
Device Adapter uses two capabilities of high availability in conjunction to minimize or eliminate single points of failure, where an IP network or server failure causes an endpoint to lose communications capability:
  • HA between a UNIStim endpoint or MGC and the Device Adapter nodes.
    HA, in this scenario, is handled by any one of the following two mechanisms. Both UNIStim phone and MGC rely on health checks for their registrations. However:
    • UNIStim phone and MGC use a form of geo-redundancy. If the primary server is unreachable, the UNIStim phone or MGC tries an alternate data center.
    • Both UNIStim phone and MGC may use an N+1 intra-cluster redundancy and this may be limited. In CS 1000, the MGC could have an active and backup call server core with which the MGC registers. However, in Device Adapter, TPS clusters scale in a manner effectively identical to the Device Adapter clusters. There could be multiple TPS node clusters with an extra “+1” node to handle large-sized CS 1000 systems.
      For more information, see HA between UNIStim endpoints or MGC of TDM endpoints and Device Adapter nodes: Intra-cluster redundancy (N+1 clusters).
  • HA between the Avaya Breeze platform cluster that has the Device Adapter nodes and other Avaya Aura® components.
    High availability between the Avaya Breeze® platform cluster and the Avaya Aura® components are handled by the Avaya Breeze® platform. This may include N+1 high availability within a Session Manager and geo-redundancy involving two Session Managers.
    For more information, see the HA and geo-redundancy between Avaya Breeze® platform clusters and Avaya Aura® components topic.
You may choose the redundancy option that best suits your requirement. For example, for a smaller CS 1000 environment migrating to the Device Adapter, it may be worth using the intra-cluster redundancy with an N+1 node environment. If you have a site with 800 devices, you may use a 1+1 environment without Geo-Redundancy, provided you are fine with outages caused because of a failure in the data center where the Device Adapters reside.
Avaya recommends that you carefully analyze the redundancy options before using them. For example, if you have two geographically separated data centers, and if the Device Adapters located at one location are out-of-contact because of a natural calamity, the Device Adapters located at other location can take over.