Open Source has its place. IBM is now a Linux shop; Wikipedia is a social Open Source phenomenon; Google Chrome just captured the majority market share of Internet browsers. When deciding between Open Source and commercial solutions, be sure you truly understand the TCO involved before making the decision.
FREE! That word has a powerful effect on decision-makers in IT Infrastructure. This allure of FREE! is the cause of more headaches, failed projects, delayed deployments, and unstable architectures than most Senior IT Managers will acknowledge. Many Senior IT decision-makers get positively feral when making purchasing decisions for Open Source solutions due to the allure of "FREE!" TCO goes out the window at that point.
Why are we so drawn to things that are "Free"? Dan Ariely, author of Predictably Irrational, believes it's due to the misconception that "free" frees us from risk of loss:
FREE! gives us such an emotional charge that we perceive what is being offered as immensely more valuable than it really is. Why? I think it's because humans are intrinsically afraid of loss. The real allure of FREE! is tied to this fear. There's no visible possibility of loss when we choose a FREE! item (it's free). But suppose we choose the item that's not free? Uh-oh, now there's a risk of having made a poor decision - the possibility of a loss. And so, given the choice, we go for what's free.
IT practitioners commonly make the mistake of going in the direction of FREE! without fully understanding the TCO involved in Open Source solutions. The two most commonly overlooked aspects of Open Source solutions are supportability and variability. Take the following two factors into consideration before you chase after your next FREE offering.
Who you gonna call? - hiring for talent and not skillsets
Supporting Open Source infrastructure solutions typically requires an in-depth knowledge of coding and programming - skills not typically found in system administrators. Out-of-the-box configurations aren't typical. I was faced with a decision of supportability recently when I had to choose between an Open Source network monitoring solution or a commercial one for a mid-size ISP. The Open Source solution came free, but it required extensive configuration and modification, as well as advanced support that involved a mid-level understanding of programming and scripting.
Good staffing practices dictate the need to hire based on talent and not skillset and to hire based on the right fit (First, Break all the Rules - Marcus Buckingham and Curt Coffman). My system administration staff had the talent and fit, but not the skillset, to support the Open Source solution, but they could easily install, configure, and support the commercial product. Time to deployment for the commercial solution < 4 hours. Time to deployment for the Open Source solution > 30 hours, and this didn't include RTFM time. At a cost per hour of resource allocation time, I saved $1300.00 in setup for a product that costs $2200.00.
We live in an age of iPods, iPads, and Xbox's - the age of GUI. Understanding technological concepts (the "talent" portion of hiring) and implementing them in the easiest way possible are the keys to running a low Total Cost of Ownership infrastructure. Having to know a particular solution's command line structure, or shell, wastes time and resources that are better allocated to productivity within the strengths of the team. My team may not know a particular Linux command and may spend 20 minutes perusing MAN to find how to do something that's extremely faster than pointing and clicking on an icon that does the job. Finding staff that understands popular commercial solutions is easier. How much easier is it to find a Windows admin than a Linux admin? Because Windows is easy enough for almost anyone to pick up - allowing me to focus on the talent and fit of the individual instead of skillsets.
101 Dalmatians - Open source is variable by nature
Open Source solutions are extremely customizable and flexible, and they typically allow access/changes to all aspects of the underlying code. These benefits are the very reason why Open Source solutions are not always the best solutions. Customizability, flexibility, and access cause variability. Variability is a blight of running a stable IT architecture.
Having begun at the very bottom of the IT Infrastructure industry, flipping burgers for a bank (also known as "helpdesk"), and working my way up through every IT department across multiple industries, I realized an important concept at play in the stable environments that was not apparent in the unstable environments - standardization. It was in 2005 that I grabbed my first copy of The Visible-Ops Handbook, thinking "Wow, How did they get inside my head like that."
The Visible-Ops Handbook analyzes what makes great IT organizations great. One of the core concepts is the idea of "cheap-prints" Infrastructure should be built on easily replicable systems that are easier to replace than to repair. Commercially available solutions reduce the complexity and variability of solutions by standardizing their solutions. This feeds the ability of creating cheap prints.
Both supportability and variability can increase costs across your infrastructure. Supportability costs are apparent; variability costs are not. Variability costs are realized through downtime for failures, upgrades, additions, and changes.
T is for total, as in Total cost of ownership. Ensuring you understand the true costs involved in supporting an IT solution can better enable you to choose correctly. Free is almost never free. The costs involved in purchasing a solution are only the start: ongoing supportability and variability add exponential costs to any solution. Be sure you are not choosing poorly and drinking from the wrong cup just because it's labeled FREE!
