Open Source. The Cheap Alternative?


One of the instinctively attractive qualities about open source software (OSS) is that it’s free. That is, you can download the software and have at it without a pesky purchase order. Depending on the license, there are no strings attached. Go ahead. Use the open source. Feel the community goodness. Feel joy at sticking it to whichever vendor you pay the largest bills to.

Only, free OSS is a myth. For an enterprise to truly leverage OSS, there is a significant investment. So, free as in puppy. Break out the newspapers.

1. Time. Operators must learn how to use the software, integrate it into the workflow, and document the appropriate processes.

2. Money. Software runs on hardware. Software must be audited. Software must be secured. Software must be maintained. Those things are quantifiable.

3. Technical debt. OSS is no different than any other sort of software. Using it creates technical debt within an organization.

Beyond these straightforward, “well of course OSS costs money” points, there are the headaches. Some OSS is powerful, but horrible to use. That could mean unstable, poorly documented, or inefficient. Or all of the above.

Cottage industries have sprung up to create a support infrastructure for popular (but difficult to implement) OSS projects. Entire companies with commercial variants of OSS exist today because they offer pain relief, as my friend Greg Ferro would say.

This is especially true of OpenStack, with companies like Mirantis making it easier to bring the orchestration platform to a production-ready state. Or BlueBox, who simply offers private cloud as a service based on OpenStack.

The company Serro recently briefed me about their OSS contributions. Their specialty? Bringing disparate OSS projects together into a harmonious whole. Define your business need, and Serro will design an architecture based on OSS project and make it all work together. They have invested the time to understand how the OSS projects work, what their programmatic interfaces are, and how to stitch them together into a solution that does what you need.

Implicit in the Serros and Mirantis’ of the world is that getting OSS right can be hard. Very hard, indeed. Therefore, the cost of acquisition isn’t the biggest issue when dealing with software. It’s the backend costs that really matter. Only when the purchase orders are compared with an eye to the future can you make an intelligent decision.

OSS might be right for you…but it’s not free. Time to put down some fresh newspapers. The puppy just made a mess.