TNS Research: How Much Did it Cost to Build the Kubernetes Ecosystem? is an article under the topic Data Science Many of you are most interested in today
Read More:
JFrog Mission Control Keeps Track of Distributed Repositories – InApps 2022
Editor’s Note: The original article has been updated to include rkt in the calculations. Also, on June 30, 2020 the interactive widget used for the calculations was replaced with screen shots.
What did the Kubernetes ecosystem cost? A couple of weeks ago the CoreOS CEO Alex Polvi posed this question. With the help of Black Duck Open Hub open source directory, we were quickly able to get an approximate answer.
Based on the lines of code in each project, we estimate that it cost about $120 million to get to Kubernetes 1.4. Albeit this is really a back-of-the-envelope guess, but it better than nothing, and it is a metric that let you compare it to other closed-source development projects.
Thought exercise, how much did it cost to get the ecosystem from scratch to k8s 1.4? Total cost of building etcd, docker, k8s itself, etc.. — Alex Polvi (@polvi) September 23, 2016
Open on Twitter.
How did we get to that number? We made many assumptions. First, that the ecosystem is only comprised of the Kubernetes, Docker, etcd, and rkt projects. Second, that the only costs associated with a projects are developer salaries. Then, Black Duck’s calculator used the COCOMO software costing model that relies on two basic inputs: 1) lines of code and 2) developer salary. Per Black Duck, “COCOMO was created to model large institutional projects, which often don’t compare well with distributed open-source projects. Beyond just development time, COCOMO is meant to include the design, specifications drafting, reviewing, and management overhead that goes along with producing quality software.”
According to the statistics on Open Hub, there are 2,609,454 lines of code in these projects. At an average salary of $55,000 a year, that totals over $40 million dollars. We know that many of Docker contributors did so in their free time, but how many is hard to determine. That being said, we also know that the average Google employee makes much more, so we looked at what total development costs would be if the average developer had salaries of $165,000 and $330,000 a year. That middle guess comes out $89 an hour per developer.
Read More:
Update Databases at Scale Part Two: The Benchmarking Game
Development Costs for Kubernetes, Docker, etcd and rkt
Estimated Average Salary | $55,000 | $165,000 | $330,000
Total Development Cost | $40,072,777 | $120,218,331 | $240,436,662
All of these figures still seem extremely low, almost chump change for those familiar with the world of enterprise IT. Don’t be fooled. The numbers do not take into consideration all the marketing, administration, risk, profit margin, et cetera that gets loaded into the final price of a project. And that doesn’t even take into consideration how much manpower it takes to install, customize and operate each deployment. That is where the real money, if not the profit, is.
CoreOS and Docker are sponsors of InApps Technology.
Feature image via Pixabay.
InApps Technology is a wholly owned subsidiary of Insight Partners, an investor in the following companies mentioned in this article: Docker.
Related Articles
Best Countries to Outsource Software Development (2026 Guide)
Software development outsourcing means hiring an external engineering team - usually based in another country- to design, build, or maintain software on your behalf, instead of hiring in-house. Most US, UK, and Australian companies do this to access skilled engineers faster and at a lower fully-loaded cost than domestic hiring allows.
How to Choose an Offshore Software Development Company in 2026
Search for "offshore software development company" and you'll get the same result every time: a list. Ten names, fifteen names, twenty-five names, each with a logo and a two-line pitch, and no real way to tell which one is actually right for your project
Are Developers Becoming Too Dependent on AI Tools?
AI coding tools went from novelty to daily habit in under two years, and the tools themselves keep getting better. But using a tool every day is not the same as trusting it, and a wave of 2026 research is starting to show a real gap between feeling faster with AI and actually being better at the job. Here is what the data says, and what it means for how you build and evaluate an engineering team.