Why Open Source Is Mainstream
August 31, 2018
This is a short piece highlighting how open source and enhanced open source software like GroundWork and others (we can use big examples like Atlassian and Cloudera) are now taken as mainstream components for many company toolsets.
There is a concept in selling products that are “sticky,” or that are not easily or cost-effectively changed out. Product categories like monitoring tend to be sticky, in that there is a significant investment in planning and implementation, in technology such as client-side agents to deploy, and significant license cost. All of these make switching to a competing product less likely.
These characteristics have been taken to extremes, and in some cases contribute to vendor lock-in, whereby the cost of using any other vendor’s product for any part of the total solution becomes prohibitive when compared to doing it the existing vendor’s “way”. In such cases companies often bend the business workflow around to fit the tool, rather than the other way around.
While it is possible to run a business this way, and to do so profitably, things have changed. Businesses are adopting cloud technologies, automation, and dynamic provisioning. In-house software development teams are working to break down barriers to change in software, creating DevOps processes rather than walled-off fiefdoms. New technologies like machine learning and AI are encroaching, and rendering obsolete (or at least non-competitive) the older software suites. Adapt or die is the rule of nature, and it is increasingly the rule of businesses in recent years.
What then is the role of open source in the context of IT systems monitoring? The reality is that most companies will not code their own software. Even if they use open source, contributing changes back upstream is the exception rather than the rule. Staffing pressures keep IT staff small, even at larger companies, so why not use supported software?
The role of pure open source at leading, innovative companies where IT is core is different from those where IT is context. Core-IT companies will develop (and often publish in open source) their own code for monitoring and management. There are many examples of this (e.g. Cloud Custodian from Capital One). Other companies will download and use this software as-is, but this practice is often frowned upon by managers, since it leaves the company dependent on unsupported code and the “tribal knowledge” of a few savvy administrators.
Several companies have instead taken the approach of integrating and supporting open source software. GroundWork in 2004 was actually one of the first of these, but now companies like Elastic, Cloudera, and others are building entire businesses around open source software, maintaining it in meaningful ways, contributing to it, and enhancing it with proprietary and open source integrations.
Complexity is your friend
This is rather counterintuitive if we apply the standards of copyright or intellectual property law to publishing of software. In most cases it makes sense for the creator of a unique work to hang on to all the rights, patents, and trademarks they can, or even to acquire them if they did not originate them. Certainly some aspects of this dynamic still apply, but the edges are blurry: true value lies more in the accumulation of knowledge and experience in a given piece of software than in the code. In most cases, code is far too complex to understand easily anyway, regardless of whether or not you have control of the source. Thus the thing that is being charged for is truly a “thing” of value: understanding of the software and how it can and should be used.
Open source is here to stay – it’s a place to start developing a solution to your problem, a place to show your brilliance to others, and a way to turn your knowledge into a business for those who can use it. It’s no longer a cult, it’s mainstream, and the pace of innovation would be far slower without it, since we would all be locked-in.