Overview of High Availability

Contents

Introduction

You can optionally deploy a pair of Edge 8000 devices for High Availability. High Availability (HA) protects SIP to SIP calls using SBC SWe Edge HA functionality. HA functionality enhances resiliency by offering Active and Standby SWe Edge instances. When the Active instance fails during a call, the Standby instance takes over the call management to maintain the call.

The Active and Standby SWe Edge instances operate using a single HA license. At the time of initial setup, you install the HA license on the Active instance. Once accepted, the HA system replicates the information to the Standby instance.

Provisioning for HA can only be done at the time of initial installation and only on a pair of devices that solely support SIP to SIP calls. Additional limitations apply. See Limitations.

For installation and HA activation instructions, refer to Workflow for Initial Setup - High Availability.

Architecture

In an Edge 8000 High Availability setup, two Edge 8000 devices share the responsibility of supporting the HA function. Each device is assigned with corresponding Ethernet interfaces to serve the requirements of the deployment, as shown in the following table:

 HA Deployment: Typical Interface Assignments

Once HA is enabled, the Standby instance monitors the Active instance through the High Availability interface. When a failover occurs, the HA function "floats" (assigns) the Ethernet 1 and Ethernet 3 addresses from the (former) Active instance to the Standby instance, and the Standby instance acts as the new Active instance. To the outside world, no change of the Edge 8000 device is discernible other than a change in the MAC address of the (new) Active device. 

The following diagram depicts the architecture of a typical HA deployment. Note that:

• Your network administrator will provide unique IP addresses for the bridge interfaces and the SWe Edge logical interfaces. (The Gigabit Ethernet ports themselves do not receive IP addresses. Rather, each port is assigned to be a member of a bridge.) 
• The physical port assignments for the LAN, management, and WAN services can be any of the available Gigabit Ethernet ports or SFP+ ports, exclusive of the port used for the High Availability interface.

HA Architecture: Deployment Example

Limitations 

The following limitations apply to Edge 8000 HA functionality:

• Deployment
  
  • HA can only be provisioned at the time of initial installation. Ribbon does not offer a path for stand-alone (SA) to HA conversion.
    
  • HA only applies to SIP to SIP calls. Analog calls are excluded.
  • HA supports only static IP addressing on the Ethernet Media (packet) and High Availability ports, not DHCP. (On the Admin port, Edge 8000 HA supports static and DHCP addressing.)
  • The devices' HA interfaces must be static IPv4 addresses on the same subnet at least three addresses apart.
  • HA supports floating IP addresses only on the Ethernet Media (packet) interfaces, not on the Admin or High Availability interfaces. (The HA function automatically assigns, or "floats," the packet interfaces of the Active device to the Standby device when a failover occurs.) 
  • HA does not support Geo Redundancy. The two HA devices must be on the same subnet.
  • Edge 8000 devices do not support redundancy at the card or port level.
• Data Replication

• • The Standby instance does not replicate Alarms, Counters, and CDR details. 
  • The Standby instance does not replicate caches.
  • The Active and the Standby instances do not store backup CDR and VQ files in the database.
  • The Active and the Standby instances do not share a Debug Facility, Monitor View, or Logging details. Each instance only shows its details.

• Call Features
  • SIP recording, Video calls, and Music On Hold do not support switchover scenarios.
• Others
  • SNMP Proxy from the Standby instance to the Active instance, so that SNMP requests always get a response regardless of which instance receives the request, is not supported.
  • Using Service Discovery to find a RAMP instance when RAMP is configured for HA is not supported.

Licenses

You must purchase an HA SKU to deploy HA on your system. Once you place an order through the Ribbon Partner Portal for HA functionality, Ribbon sends you an email that contains information on where to obtain the license associated with your system. You will receive one license for the HA pair.