Open Source Communications for the EnterpriseOpen Source Communications for the Enterprise

With Asterisk Scalable Communications Framework (SCF), Digium and its affiliated community have a new architecture aimed at large enterprises and carriers.

The notion that Asterisk, the open-source PBX, was not for the enterprise was one of those "facts" that everybody knew, or "knew." Digium, chief sponsor of Asterisk, is able to cite some big deployments or intended deployments--a university consortium in Portugal, some municipal governments--but large enterprises basically haven't been willing to take the plunge.

But does that mean enterprises won't go for open source communications as a matter of policy or technology religion? Digium has just placed a big bet that says open source communications can make it in the enterprise, and even in the carrier space.

That bet comes in the form of Asterisk Scalable Communications Framework (SCF), and for a great overview, you need to read Dave Michels' SCF piece at UCStrategies, as well as a companion posting at Dave's own Pin Drop Soup blog. There's also an extensive Asterisk SCF FAQ on the project wiki that Digium is running, as well as the main Asterisk SCF page

The key word in the product's name is "scalable," which of course is a prerequisite for any system to be considered by large enterprises and carriers. SCF is an entirely new code base, with essentially nothing in common with Asterisk. At Astricon last week, I had a chance to sit down with Danny Windham, Digium's CEO; Global Marketing VP Leslie Conway; and Senior Software Engineer Kevin Fleming, and they described how the project came about and where Digium intends to steer it.

Late last year, the top execs at Digium came to the conclusion that Asterisk was at a crossroads, and so that November, Kevin Fleming and other technology leaders held a day-long meeting, at the end of which they concluded that Digium needed to make a "clean break" from the original Asterisk and build and entirely new architectural framework. The path from that meeting to, almost a year later, announcing SCF at Astricon, involved getting enough of the framework in place and sketched out so that the Asterisk developer community would have something to sink their teeth into, without presenting them with a fait accompli, which would defeat the purpose of an open source project.

From Danny Windham's perspective, Digium's overriding concern was to keep Asterisk in its position as the most successful open source communications system in the world; ultimately, that meant re-focusing the Asterisk developer community on a totally new project.

In his Astricon keynote announcing SCF, Kevin Fleming demonstrated a simple failover scenario, something that Digium will be stressing, because, as Kevin says, the distributed architecture lends itself not only to scalability but to resiliency--the technology will be attractive for "really really large deployments, especially those with stringent uptime needs," he said.

The failover demo from the Astricon keynote is basically all there is to show right now, Kevin said; he told me that he realistically expects beta-quality code to be available in about a year.

So is Asterisk SCF a big deal for the enterprise? I'd argue that it is--or at least that it could be. The first-generation Asterisk may have had the limitations you'd expect from a system that one guy wrote in his spare time to serve a very small implementation of his own. Lots of enterprises seemed to view Asterisk as a lab project, but fundamentally, open source is a perfectly respectable approach to technology development, one that probably has more acceptance in other parts of the technology organizations within the enterprise, outside of communications/IT. A scalable open source communications system--if it's proven--would not necessarily be a non-starter at a lot of enterprises.

On the other hand, Digium is in something of a race against time, it seems to me. In a couple of weeks, Microsoft is going to officially launch Lync, the third generation of OCS, and it's going to get a lot of attention from enterprises. Cisco has a toehold from both the data network and the legacy IP PBX market. Avaya is doing its best to hang tough. With pent-up demand finally being somewhat released, and some purchases, at least, going forward, a lot of the future of the market may be shaped in the course of the next year.

But if nothing else, Digium has absolutely got a differentiator that enterprise managers can grasp immediately, and a track record with Asterisk that shows they can foster a community of developers and channel partners that can develop and implement a product that's effective for its market. In the case of the original Asterisk software, that market skewed very heavily to the low-end of the market, but if SCF delivers the goods on scalability and flexibility, enterprise managers will have a legitimate option should they be open (so to speak) to it.

Making the clean break via the new SCF will cost Digium some time on the front end, but they can argue that it's a more forward-looking platform on which to build your communications. Asterisk SCF is a significant development in enterprise communications.