svi functionality

High Availability (Dual Redundancy)

Overview

• Within any C21st telecom network, customers expect the service they are using to be available 24 hours a day, 365 days a year.  With many companies providing similar services it becomes imperative that customer churn is reduced by retaining as high as possible availability of the service. 
• 
  To increase the availability of any service it is important to identify any single point of failure within a network and the network's equipment, so if failure does occur there is no service outage.  High Availability equipment is specially designed to eleminate single point of failures to ensure the continuation of service in the event that a hardware or software resource fails with the system.

Technical Overview

• The most common and effective way to provide high availability is to provide a dual plane architecture.  This generally involves two planes or servers that replicate the exact environment on both planes but appear to the network as a single entity i.e. providing a single IP address or SS7 point code into the relevant network.
• One plane acts as the worker plane and the other is the standby plane.  The worker plane is the live unit that handles all of the call information gathered from any signalling from either the worker or standby unit.  The worker plane then updates the standby plane with any call change information or updated local database information.  If any resource on the worker unit fails the standby unit is in a correct state to take over the processing of the call information without affecting the service.
• To maintain the service the signalling transport must be connected and operational to both planes for the system to be fully redundant.  The following shows how this is achieved with each of the different external interconnects.

Squire Solution

• SS7 TDM Interconnect

• Figure 1 shows a dual plane system connecting to a SS7 TDM interconnect.  Because SS7 provides a redundant signalling architecture the signalling links can be split across the two planes.  Because the hot standby terminates the SS7 signalling at MTP L2 it can transport the SS7 signalling it receives across to the worker plane therefore providing the ability to maintain a single point code across the interconnect as only the worker plane handles any signalling above MTP L2.  If a failure occurs at the TDM level or on a single plane, signalling traffic will still be able to be sent on the other plane with the
• effect being that the total signalling bandwidth is halved.

• SIGTRAN M2UA Interconnect
• 

• Figure 2 shows a SIGTRAN M2UA interconnect.  This works in a very similar way to the SS7 TDM interconnect with the MTP L2 being transported over IP instead of TDM.  The added advantage of using M2UA is that a redundant SCTP transport layer can be setup to the signalling gateway, as indicated by the dashed lines, which provide extra redundancy for a single M2UA signalling link in the event that a M2UA's IP transport is lost allowing a different path for this data.

• VoIP Interconnect
• 

• To provide a single point of interconnect into a VoIP network the SVI uses the concept of a VIP (Virtual IP) address.  Each plane has a dedicated IP address as shown in Figure 3 with the worker plane having an IP address of 192.168.2.200 and the Hot Standby having an address of 192.168.2.201.  The current worker plane also has a second IP addresscalled the VIP address which in Figure 3 is 192.168.2.211.  Any VoIP service is always directed at this VIP address.  On a failover the VIP address will be removed from the old worker and allocated to the new worker therefore any signalling communication will be automatically directed at the new worker plane unknown to the interconnecting VoIP equipment.  If the VoIP signalling is using UDP (SIP) as the transport layer then no service is lost, if TCP (H323) is used then existing calls will be lost on the failover.  The VoIP media path is not affected on a failover because the media path is transmitted either directly between the VoIP end points or between the VoIP end points and the media gateway.

• Media Gateway Interconnect
• 

• The SVI when operating as a Media Gateway Controller communicates with the Media Gateway via either the MGCP or H.248 (MEGACO) protocol.  Both of these protocols run over UDP, so the concept of the VIP address is used again to retain a single IP service address into the Worker/Standby pair.
• All of Squire SVI range has been designed to operate in a dual redundant, high availability environment providing up to 99.999% uptime.