TMCnet
TMC Launches New Sites ::  4GWE  |  Satellite  |  IT |  Business Video  |  Healthcare  |  Smart Grid  |  M2M  |  Smart Products  | 
Share
Asterisk

Resource Center

Featured Podcasts
Leveraging Asterisk for Your On-Premises IP PBX: Erik Linask talks with Digium’s Tristan Degenhardt about the IP PBX market and the latest enhancements to Digium’s Switchvox product.
Rich interviews Danny Windham,
CEO Digium

Asterisk White Papers

Switchvox IP PBX - The next generation of business phone systems
Digium Product Spotlight
Asterisk Appliance
* Full-featured, cost-effective
Internet Telephony in a User-friendly Package.
Asterisk Developer Blog
Discover Asterisk 1.4 :: Jitterbuffer, Jabber and blinking lamps I've written a series of articles covering new features in Asterisk 1.4, the one-year old release that is finally getting stable enough so I can recommend it for production use to my customers. It's high time to discover 1.4 and start ...
Webinars

Asterisk Case Studies

Asterisk Featured Article


Talking with Mark Warren, CEO of Redfone Communications


 Mark Warren is the CEO of Redfone Communications, LLC (www.red-fone.com), a company that has brought high availability to the world of Asterisk (News - Alert) -based, Linux-powered, open source Voice-over-IP. Prior to Redfone, Warren was Technology Manager at the Sabre Travel Network and Sabre Holdings, from 1997 to 2005. Previously, he ran International Operations for the TV Guide Channel (1995 to 1997). He had worked for the Prevue Channel, which was acquired by TV Guide.

 
RG: Admittedly, just using the terms “open source telephony” and “fault tolerance” in the same sentence is pretty mind-boggling. Why did you bestow such technology on the humble world of open source?
 
MW: We – Redfone’s founders – had used Digium’s (News - Alert) Asterisk-based VoIP technology in our previous avionics business. Surely anything so  inexpensive with a good IP PBX feature set has a great future, particularly for small and medium-sized businesses. However, what it lacked compared to expensive communications systems was true fault tolerance, which would give it equivalent reliability and hence, credibility, and enable open source deployments in more demanding environments.
 
Asterisk software generally runs on a single server, with PCI-based T1 or E1 interface cards connecting the system to the outside world. Such a configuration, of course, is a single point of failure and puts voice traffic at risk. Voice is still the true “killer app” and I can’t think of very many companies that would feel comfortable with having their phone calls continually in peril.
 
RG: So you devised your own hardware solution?
 
MW: Our second-generation hardware is an appliance-like external device called foneBRIDGE2. It can be plugged into the network by IT managers and systems integrators to handle Linux clusters running Asterisk and other open source software-based telephony platforms, such as Fonality’s (News - Alert) trixbox.
 
Since we rely on Ethernet for connectivity between servers, the system’s scalability factor increases dramatically, because hardware alterations and software upgrades performed by having a server admin manually take down a server and thus cause the service to failover to one or more other servers comprising the Linux cluster, which automatically take up the slack. Of course, with foneBRIDGE2 you configure multiple Asterisk servers that don’t actually have any TDM (PRI/T1) hardware cards in them. Instead, foneBRIDGE terminates to PRI or T1 channel banks on the trunk side and Ethernet connects to the Asterisk boxes.
 
Similarly, any unintentional event or problem that takes down a server also triggers rapid failover on the cluster. foneBRIDGE2’s rapid failover and system recovery occurs in seconds. The foneBRIDGE2 re-configuration options work very quickly, and you can re-program it “on-the-fly” via high availability tools so that a TDM phone call stream can be routed to a secondary standby server occur in under one second.
 
Therefore, imparting high availability principles to a server environment doesn’t just guarantee failover and continued VoIP service, but it also provides you with the related benefits of increased serviceability and scalability.
 
RG: What kind of software makes this possible?
 
MW: One popular HA toolset that works well with both the foneBRIDGE2 and Asterisk is the Linux-HA project’s ‘heartbeat’ tool (www.linux-ha.org). The use of a “heartbeat” in the world of high availability has a long history. Basically, messages – the ‘heartbeat” – are sent to the server nodes in the cluster. If a particular node doesn’t respond to the heartbeat messages, scripts are executed to quickly failover the voice service to another server. This process is completely automatic and transparent.
 
The foneBRIDGE2 also supports “mixed-mode” operation support, so that T1, E1 and RBS/CAS services are configurable on a per-port basis. Interestingly, for all of its features, foneBRIDGE2’s price points compare favorably with to competing dual and quad-port PCI products.
 
RG: How easy is the whole thing to install? What if a customer has questions?
 
MW: We at Redfone Communications support both Asterisk and trixbox users. We can handle everything from basic questions and help with configuring the system, to more elaborate engineering consultations concerning how to correctly dimension and generally implement your system. Thanks to our efforts in developing foneBRIDGE2, even small businesses can now enjoy the same reliability, scalability and feature set as can be found in a very expensive PBX (News - Alert)-type phone system. The open source telephony model is now looking better than ever.
 

Richard Grigonis is Executive Editor of TMC (News - Alert)�s IP Communications Group. To read more of Richard�s articles, please visit his columnist page.

Edited by Tim Gray

 

Participate in the Community:
Add Your Thoughts and Comments Now
 
 
By  
TMCnet
 
Follow the Asterisk Community on Twitter