KubeCon China – Cloud Native Computing Foundation (CNCF) celebrates 10 years of growth. But are clouds gathering?

KubeCon + CloudNativeCon China 2025 is a leaner and more focused event than last year's conference in Hong Kong. In 2024, the Cloud Native Computing Foundation (CNCF) hosted a three-day jamboree, which incorporated the Open Source Summit and AI_dev China, featuring luminaries such as Linux prime mover, Linus Torvalds – see my reports from last year.

In 2025, the conference has shifted from the Kowloon side of the water to the center of Hong Kong Island, where it offers a more concise two-day program. More significantly, Artificial Intelligence (AI) no longer requires its own track or summit; it is now part of the everyday reality of cloud-native and Open Source computing. Last year? Well, that was a century ago in AI development.

So, what of the CNCF itself, now that AI has moved center stage, just as KubeCon + CloudNativeCon China has moved to the center of Hong Kong? A city that, in the real world, finds itself sitting beneath storm clouds: we are early in the typhoon season. Perhaps those clouds are another apt metaphor? More on that in a moment.

But first the good news. Chris Aniszczyk is CNCF Chief Technology Officer. He explains that 2025 is a landmark year for the organization:

It marks 10 years of CNCF. And since we started, we've had an incredible amount of growth. We're one of the largest Open Source organizations out there in terms of global contributors. We're at about 287,000 contributors worldwide across all our projects [up from 270,000 in April]. And there are well over 200 projects now in the organization spanning contributors from over 190 countries.

Impressive stuff. And in addition, there are 728 Member organizations and 100,000 community group members, with other landmark figures from the past 10 years including nearly 19 million code contributions and 1.2 million pull requests.

As Linux Foundation Director Jim Zemlin put it on day one of this year's conference, "Open Source really works when you build a community around it." And the CNCF has certainly done that alongside the Foundation.

It is showing real results too. CNCF research published in April found that there are 9.2 million cloud-native developers worldwide, up 2.7 million from 2020, with development taking hold in organizations of every size: adoption stands at 89% overall.

Breaking down that figure, 61% of organizations say that much or all their development is now cloud native – that includes 67% of small enterprises and 59% of large ones (those with more than 5,000 employees). Of the remainder, 28% say that at least some of their development is cloud native.

The cloud figures themselves are interesting: 39% are using a hybrid public/private environment, 46% a managed public cloud, 59% a self-managed public cloud, 40% an on-premise private cloud, and 59% a managed on-premise deployment. Those figures illustrate respondents' multiple cloud environments, hence them exceeding 100% in total (respondents could select more than one option).

Rising complexity concerns despite growth

But not everything is rosy in cloud-native's Hong Kong garden: the survey finds 46% of respondents saying that projects can be too complex to understand and run – up from 33% in 2023. Nearly as many (45%) say that supporting documentation is thin on the ground – up from 40% in 2023 – while 27% say that Open Source projects can be difficult to install, up from 25% in the previous year.

The better news is that users believe the environment is becoming more secure: 29% report vulnerabilities, down from 36% in the previous year. That said, 46% of respondents noted a worry that Open Source projects could become inactive, up from 37% in 2023. Interesting, if you read between the lines.

So, who is an Open Source user? According to CNCF, 38% of them are based in Europe, including Russia, 37% in North America, six percent in India, and just three percent in China – intriguing, given that China is the second largest contributor to projects worldwide (who do they all work for?).

Meanwhile, 38% of all users are in small organizations, and 36% in large. By far the largest user group comprises software designers and engineers (45% of respondents), and 61% of users have more than 10 years' experience. All of which brings us to those storm clouds I mentioned.

I may be wrong, but the above figures might suggest problems ahead for the Open Source community in the AI age, in which more and more coding is being handed to machines. If the median Open Source proponent is an ageing (in software terms) European-based programmer, who is worried about rising complexity and inactive projects, then the future may not be quite as bright as the community claims.

Put simply, despite the youthful buzz of the KubeCons I have attended in China and Paris in recent years, we might be looking at the old guard, not the new, in development terms: experienced coders who are facing a possible future of rising automation and AI – aka vibe coding. But only if some noisy AI CEOs are to be believed: the kind whose investors love to believe that all essential work can be handed to machines without a second thought.

But what of China itself? Aniszczyk explains:

As you know, China is already the number-two contributing country to CNCF projects, a great strong signal. […] There has been a lot of innovation born out of China in particular, with a lot of great, large-scale deployments here. And a couple of handfuls of CNCF projects were born here, or mostly in China.

We have [peer-to-peer file distribution and image acceleration system] Dragonfly; Volcano, one of the early batch scheduling projects from CNCF, that's used widely in the AI space. KubeEdge, obviously for edge computing, which graduated not long ago, and a lot of other efforts from China shared to CNCF and shared with the world.

So, where does AI fit into this picture?

Lin Sun is Head of Open Source for cloud-native API management platform solo.io. Speaking on day two of KubeCon in Hong Kong, she explains:

What I believe is that 10 years ago, Kubernetes was the change agent for cloud native. But today, AI is the change agent.

Remember: the event is KubeCon; ostensibly, a Kubernetes conference.

She continues:

Now AI and agents [agentic AI] are forcing us to rethink compute, network, and storage in Kubernetes and cloud-native projects. And AI is also challenging us to rethink some of the fundamental building blocks in Kubernetes, such as Pods and services.

Doubtless that is true, hence the rising fears among developers about complexity and dormant projects, perhaps. But that is only part of the picture, she explains:

What I believe is the most complex ecosystem is not about AI, it's about the complexity of the mobile ecosystem. And you have been seeing the very complex cloud-native landscape we have. So, any new user who comes to cloud native has to kind of suffer the pain, the complexity, of learning cloud-native projects.

And I believe the next wave of AI is going to be around how we bring a simplified user experience to cognitive users, so we can reach out to the next 10 million, the next 100 million users of cloud native.

My take

Indeed. And how the community does that would seem to be the billion-dollar question.