Nov 14 2008
High-Availability/Failover Asterisk clusters
If you use PRIs and are looking to create a reliable, fault tolerant, highly available system with Asterisk, the foneBRIDGE by Redfone is a solution worth looking into. The foneBRIDGE provides quick, reliable and robust T1/E1 fail-over allowing IT Managers and System Integrators to implement Asterisk in the most demanding of environments where downtime simply can not be tolerated. The foneBRIDGE is now available in a 1U Rack Mountable Enclosure that supports upto 8 T1/E1 interfaces.
High Availability
The general objective of deploying HA (High-Availability) in a server environment is to provide reliability, availability, and serviceability (RAS). In the traditional Asterisk implementation model, T1/E1 connectivity is provided by PCI based interface cards on a single server. In this scenario the systems overall integrity and availability is limited by the reliability of that single server. For many businesses, taking this gamble on a mission critical function such as voice is unacceptable.
As an externalized point of T1/E1 termination, the foneBRIDGE is decoupled from the single server limitation and when combined with the proper tool set can be implemented to provide rapid and automatic failover amongst a cluster of servers.
The most common tool set used to provide this functionality when paired with the foneBRIDGE and Asterisk is the ‘heartbeat’ tool from the Linux-HA project www.linux-ha.org. As the name employs, ‘heartbeat’ monitors the state of 2 or more nodes (servers) in an Asterisk cluster. If it detects a fault, for example when a primary node stops responding to the heartbeat messages it executes scripts that allow it to quickly failover to a secondary server. All of this is done transparently and without user intervention.
Rapid Recovery
With the foneBRIDGE, rapid failover and system recovery can occur in seconds, not minutes or hours. Through the foneBRIDGEs’ rapid re-configuration options, it can be programed on the fly through high availability tools to begin routing the TDM stream and calls to a secondary standby server in under one second. Add to that the short amount of time it takes the server to start Asterisk and clear circuit alarms and your stand-by server can be operational and handling calls in under 3 seconds, all transparently and without administrator intervention.
Uptime During Maintenance and Upgrades
Over are the days when it was necessary to bring down a server for patching or software upgrades. With foneBRIDGE, the server admin can manually failover to a back-up server and maintain the voice system operational while performing upgrades or patching on the primary server. This unique feature also lends itself nicely to the ability to failover to a server where production testing can be done on new code releases or features and in the event of a bug discovery or issue that may jeopardize the integrity of operations the system admin can easily fail back to a primary server with only seconds of downtime.
Scalability
With connectivity externalized, and placed on the Ethernet not tied to any PCI card or single server, adding additional T1/E1 capacity to an Asterisk implementation is simplified and can be realized with little or no downtime to operations.
For more information, visit the foneBRIDGE product page
