Five Tenets of Thriving with Open Source without Risking Your Business
1. Open source is not free.
Vendors, university researchers, students, and developers find that open source is a very effective tool for validating a solution to a particular problem. However, these solutions are often born within the context of a specific company’s use case or a specific research problem. Therefore, these projects are similar to code built in the context of a professional services engagement and often don’t have the polish and finish of an enterprise product. Some may not even have solved basic enterprise requirements like availability, resilience, security, and so on.
Therefore, adopting open source in the wild comes with its own risks and resource investments.
2. Learn to identify the enduring open source projects and stay away from the flash-in-the pan open source projects.
Just like startups, many open source projects are born, but only a few popular ones withstand the test of time. Another analogy that is apt is a mobile game—many games are created, a few are popular for a brief amount of time, but only a handful have endurance.
In today’s world, there seems to be a new open source project every week that can keep organizations distracted. Pick partners who continuously assess the tech landscape, but are focused on helping you achieve your business goals; they will save you the heartburn of many fruitless endeavors. Make no mistake, some experiments will still fail, and those that succeed will have a more strategic and lasting impact to your business.
Apache Hadoop and Apache Spark are examples of projects that have demonstrated the endurance in the broader big data analytics ecosystem. Apache Kafka’s publish/subscribe API is another one that has shown the promise to solve a large problem in the enterprise data architecture. Open source NoSQL databases also helped move applications away from the classic rigid schema-based RDBMS model and ushered in a new generation of applications. However, many other compute and storage projects have either failed or found narrow adoption in select environments.
There are extremely promising trends in compute, storage, and networking that foretell a brave new world for enterprises
Here are a few questions to ask to help identify enduring open source projects. This is certainly by no means a complete and exhaustive list.
• Is it an Apache project?
• Does it solve a real significant problem or does it sound like a feature your product should have?
• Is the focus of the project on use cases or on technology jargon?
• Is the project backed by a solid team of engineers?
• Do the engineers have the financial backing and capacity to push through the productization and generalization phase of open source?
• Is it the right technical approach to the problem?
• Is there a diversity of interests in the community team?
If the answer to many of these questions is no, it is highly likely that the project will either serve the interests of a few narrow companies or will be short-lived.
3. Open standards are more important than open source.
In the last few years, we have observed an unprecedented amount of dollars and energy poured into marketing open source software companies, and the general mantra has been that “open source is always better”—because it will evolve and eventually catch up with better systems or will make it easier to hire the right skill set. Failing fast and moving from one open source technology to another is very hard, as each open source project has its own API (e.g., the NoSQL open source space). Therefore, enterprises get stuck with the same problems that they had with closed source systems.
As the dust settles, we observe that enterprises are now wiser and more nuanced about open source. The transition is towards a multi-dimensional rigorous criteria defined around open standards, open community, project endurance, and most importantly, business relevance.
Open source continues to thrive as a platform for experimentation with new ideas and technologies. It is the mainstay for research students to explore new ideas by tweaking publicly available systems and putting their new theories and algorithms to test.
4. Converging operations and analytics is the key to digital transformation.
We are living in the “golden age” of technology. There are extremely promising trends in compute, storage, and networking that foretell a brave new world for enterprises.
However, the ability to digitally transform an organization requires us to fundamentally accept that operations and analytics have to work together in real time. Insights from analytics have to feed operational decisions and observations from operational execution have to feed analytics—all in real time. This is the key that has helped Airbnb become one of the most valuable hospitality companies and Uber to become one of the most valuable transportation companies.
5. The future is bright.
With the right technology processes and always available data, experimentation can be inserted into the fabric of the enterprise IT architecture instead of a separate initiative with the best and the brightest working on it. Assessing, experimenting, and moving new technologies into production to realistically assess their benefits and reliability can be done without jeopardizing live production processes and the heavy cost of procurement, setup, and POCs.