Open source calling

I am running out of options for areas in my IT operation that legitimately shouldn’t be open source. Operating system? Linux works like a champ. Web server? If you’re not running Apache at this point, what are you doing? Database layer? MySQL scales fine for most Web-based apps, and basic master/slave software clustering for it is free, which can save roughly six figures over a commercial solution if you’re running more than a couple of database servers. App server? JBoss if you want Java, or you could just use PHP running on Apache, among many other choices. OK, I haven’t spent any money on software yet, and hardware is cheap. I’m surveying my office right now, looking for something that I couldn’t enable with open source software, and my eyes fix on that ugly corporate phone that hooks into the old PBX. I feel helpless before it -- I look at it and the words “lock in” might as well be blaring from the speakerphone. There’s nothing I can do about it. Open source can’t help me with my crusty old PBX. Except that it can. And for me, that suggests that open source can -- and will -- go anywhere and everywhere.

Free IT resource Virtualization Insights from Top Experts - Learn how virtualization gets real! Sponsored by Dell Free IT resource TechNet: More ways to know it, share it, and keep it running. Sponsored by Microsoft

Asterisk, a highly sophisticated open source PBX software package that runs on commodity hardware running Linux or pretty much any other operating system with GCC (GNU Compiler Collection), is the first open source project in quite a while to really make me stop and pause. The InfoWorld Test Center gave Asterisk high marks earlier this year and O’Reilly’s Nat Torkington recently noted on the O’Reilly Radar blog that his company’s IS department planned to replace its production PBX with Asterisk. I shouldn’t be surprised because our reviewer gave Asterisk such strong praise -- and on low-end desktop machines, no less. Instead of fearing such a critical application of open source, I wish that I ran the phone system here at InfoWorld so I could start planning to roll it out myself. Assuming reasonable planning, I fully expect the O’Reilly implementation of Asterisk to be as successful as our own tests were.

For the doubters who might say, “I would never use open source for something as critical as my phone system,” I offer some historical perspective. Six years ago, switching to Linux was considered daring enough to garner me a mention on Slashdot and an interview in PC World. Now Linux is so routine no one really cares anymore. I chose Linux then because it worked flawlessly, just as it does now in my current environment. Looking farther up the stack, remember when IT used to say, “We can run our Web servers on Linux, but when it comes to my database, I’m sticking with the big guns?” A quick glance at MySQL AB’s current customer list (France Telecom, Google, Suzuki, to name a few) suggests that the once-revered database layer is no longer sacred. No doubt we will all look back five years from now and feel the same way about open source telephony solutions as solutions such as Asterisk slide down to the comfortable end of the fear curve -- just as Apache, Linux, and MySQL have done before.

For IT managers, the question becomes, “Who is going to fix this stuff when it breaks?” Probably the same smart people who run your Linux/Apache Web servers now, only they will handle hardware and software upgrades on their own instead of relying on expensive engagements with old-school telephony vendors. Think this is crazy talk? Give me a call in 2010 -- and let’s see what type of PBX is handling the connection.