The central theses
- The latest CNCF survey claims that 5.6 million developers are using Kubernetes in some way or form.
- Based on the 23% of developers worldwide using it, Kubernetes has crossed the divide into the mainstream early majority category.
- With 39% of respondents implementing serverless in one way or another, this is clearly the year of serverless…as it has been for more than 5 years.
- This community has a healthy software release cycle, with approximately 50% of respondents reporting that they release code to production at least monthly, 31% do so weekly, and 18% daily.
- Although Amazon Web Services (AWS) was an early competitive motive for Kubernete development, it is the top choice for hosted Kubernetes. Kubernetes is truly an “everyone wins” game.
You know I love good surveys, so let’s take a look at the Cloud Native Computing Foundation’s 2021 annual survey. They asked 2,302 respondents how they use Kubernetes and the broader category of cloud-native tools. The main conclusion of the report is that the use of Kubernetes is mainstream, as the report’s subtitle indicates 2021: “The year that Kubernetes crossed the abyss.”
This year’s survey has numerous topics, but let’s look at the three topics that interest me the most.
When the CNCF survey has the tagline “the year Kubernetes crossed the chasm,” it means “Kubernetes breaks into the mainstream.” Let’s verify this claim. First: “Abyss”?
Concerning the theory of innovation diffusion, Geoffrey Moore’s work Crossing the Chasm posits that the introduction of software to the market goes through five phases: innovations, early adopters, the early majority, the late majority, and the laggards. Each group is increasingly less tolerant of new, more difficult-to-use software. Innovators will do anything because they like new technology, laggards may not even care about your new software. There’s a gap between visionaries and the early majority that most software companies can’t bridge. This is where what worked for one provider (now also public cloud) with this small group of customers no longer works as well for everyone else. They must “scale” their product management, engineering, support, sales, and marketing to expand into organizations that have increasingly mundane, normal, even boring causes. So Moore’s model is good for finding out if and when Kubernetes has become mainstream.
Then let’s do the math to see if Kubernetes made it across the abyss. Each group in Moore’s model shows up (surprise!) in a bell curve with innovators representing 2.5%, early adopters 13.5%, early majority 34%, late majority 34%, and latecomers (fortunately!) only 16% . So to cross the abyss you would need 16% of the people using your software. If you’re in the 17% to 50% range, you’re in mainstream territory.
Now let’s calculate how many global developers are using Kubernetes. SlashData, which conducted this survey for the CNCF, estimated that there were 26.8 million developers worldwide in 2021. The CNCF survey states that “Today, 5.6 million developers are using Kubernetes.” So we get the following pie chart:
If my calculations are correct, 21% is greater than 16%, so Kubernetes is bringing the abyss. Now, a word of caution: I don’t know how SlashData and the CNCF came up with this estimate of 5.6 million developers using Kubernetes. But let’s continue.
I think we all intuitively know that Kubernetes is just beginning its mainstream life, and it’s always good to have some numbers like this to back it up.
However, with something like Kubernetes, I like to keep an eye on the proportion of all applications: Of all the applications running in the world, how many are running on Kubernetes? So far, this pie chart has eluded me. Some analysts have tried, but I don’t think we have a handle on Kubernete’s share of all applications yet. Or maybe the numbers exist and I just need to renew some of my analyst seats and brood over their delicious pies. Hopefully the next poll will address the question of usage share.
2021 is the year of serverless… again
With 39% of respondents going serverless, this is clearly the year of the serverless…but that’s been the proclamation for the last 5+ years. Sarcasm aside, growth, while small, is accelerating significantly. Serverless is definitely “a thing”, but it’s been used with the same overall usage rate for a number of years.
Since 2016, if not sooner, serverless was on the verge of changing everything every year. Will 2022 be the year of serverless? In the 2021 survey, 39% of respondents reported using serverless technology. Let’s see how this compares to previous years in the table below.
It’s clear that something strange has happened in 2020, but the overall trend is the same. Small changes in usage over the years might lead you to believe that all the people who are going to use serverless are already using it. Additionally, O’Reilly reported a sharp decline in serverless training over the past year. Or you could be more optimistic: if you look at the evaluation and planning responses over time, you could theorize that there’s a second wave of people interested in serverless but not using it yet. Digging this story further, one could say that this cluster is the late majority and laggards. However, since these numbers are so varied, I wouldn’t be too specific in my conclusions. We’ll have to wait and see what the next few years will bring.
While we wait, let’s look at the serverless public cloud options that users are using for serverless applications. Amazon, Azure and Google top the list for hosted serverless services. This has been the case in the polls since 2018. Looking at “installable software” for serverless over time is a lot more fun because you see a lot of comings and goings (keep your margarita for this time):
In this chart, I’ve only included the top three each year (except 2021) and rolled them over to subsequent years to see how they’re moving. As a result, many options were missing, for example 15 frameworks were available in 2021, including “Other”.
What you see over the years using serverless frameworks is that OpenFaaS has caught on and Knative has quickly gained popularity. That said, I wouldn’t start breaking champagne bottles just yet: In the 2021 survey, only 151 people answered this framework question, with 1,376 people skipping the question. There are still many people considering serverless.
Time-to-market check: software release cycle
I’m always interested in how long companies need to get from the idea to the software in production. The more frequently you can release software, the more feedback you’ll get about the usefulness of the software because you can better monitor users and adjust features accordingly. There may be applications that are exceptions, but I haven’t seen them yet. How is the community doing with software releases?
While it’s dangerous to establish causality between Kubernetes usage and software release cycles, people who took part in this survey do well when it comes to frequent deployments. In this year’s survey, over 80% of respondents said they release code to production at least monthly, with 31% doing so weekly and 33% daily.
Although incremental, there is a major trend towards shortening release cycles over these four years. This is key to improving the design and effectiveness of your software because you can set up a small batch loop that allows you to experiment with new features each week and learn what works and what doesn’t, so you keep making your software better and better do.
Summary: This is not a zero-sum game
After looking at the current Kubernetes documentation – which is great, you should check it out – it’s worth reflecting on the result of Amazon Web Services topping the list of public cloud Kubernetes services, as well as some of the others Questions and answers. In the Kubernetes documentation, you’ll hear that Kubernetes was originally a competing answer to Amazon, among many other motivations. Not only did Amazon eventually join the Kubernetes party, but it has grown to become the most widely used distribution and hosted Kubernetes service. This reflects an important point of the documentary, which is that Kubernetes is not a zero-sum game: everyone “wins” interest, especially users.
There’s a lot more in this year’s survey, and certainly more in the raw data if that’s something you’re interested in. And if you’re interested in something like this, also check out my report on our 2021 State of Kubernetes survey.