James Governor's Monkchips

Rise of The Docker Pattern

Share via Twitter Share via Facebook Share via Linkedin Share via Reddit

Sharp Green by Debbie Clapper

Once you’ve been in the industry for a while the patterns become clearer. Enterprise technology adoption has some fairly distinct shapes. In 2009 I wrote

“Amazon is the new VMware. The adoption patterns are going to be similar. Enterprises will see AWS as a test and development environment first, but over time production workloads will migrate there.”

I dubbed this “the VMware pattern”. New technologies generally don’t emerge as fully-fledged production environments. They are adopted first and grow into the role. Docker is currently on a fast track through this process.

In the same post I wrote:

“Amazon isn’t the de facto standard cloud services provider because it is complex – it is the leader because the company understands simplicity at a deep level, and minimum progress to declare victory.”

For docker replace “simplicity” with “convenience”. Why is Docker is so hot? The answer is simple. Developer-led adoption, or as Andrew Clay Shafer puts it:

“It’s the fastest path to developer dopamine”.

At RedMonk we have never seen a technology become ubiquitous so quickly. Docker makes it simple to spin up a container which contains everything needed to run an app – the code itself, the runtimes, systems tools etc. Develop on your laptop, then in theory deploy to any server. Unlike virtual machines, containers include the application and all of its dependencies, but share the kernel with other containers, an efficient model which maps cleanly to current development thinking in areas such as continuous integration and microservices. Stephen, in a thoughtful explanation of the Docker phenomenon, argues that:

The explosion of Docker’s popularity begs a more fundamental question: what is the atomic unit of infrastructure moving forward? At one point in time, this was a server: applications were conceived of, and deployed to, a given physical machine. More recently, the base element of an infrastructure was a virtual recreation of that physical machine. Whether you defined that as Amazon did or VMware might was less important than the idea that an image resembling a server, from virtualized hardware and networking interfaces to a full instance of an operating system, was the base unit from which everything else was composed.

Containers generally and Docker specifically challenge that notion, treating the operating system and everything beneath as a shared substrate, a universal foundation that’s not much more interesting the raised floor of a datacenter. For containers, the base unit of construction is the application. That’s the only real unique element.

So alongside many of the other micro and macro trends we currently see, notably infrastructure fragmentation, Docker basically just makes sense – it feels right and represents how developers live now. Next however comes the fun part – Docker will begin to reshape how operations and IT work, just as VMware did in the virtualisation wave.

Enterprises needs to find ways to deliver more digital services to market faster, which means not only becoming more adept at developing, but also consuming new technology. Docker can help with that. In the age of continuous deployment dev/test underpins the deployment process. Everything needs to be constantly tested, and constantly refactored, with an eye to disposability rather than reuse. See Microservices and Disposability: On Cattle, Pets, Prize Bulls, Wildebeests and Crocodiles

Docker is on an exceedingly well-funded mission to transform itself from developer favourite to Cloud Native production environment of choice for the enterprise, moving from Open Container format to “single virtual computer” of choice. The transition though from developer-led to enterprise production grade takes time. We’ve seen this before, from MySQL to Mongo to Spring… or for those with rather longer memories think the early versions of Oracle. Automation, backup, compliance, logging, monitoring, networking, scheduling, storage management, orchestration, security, and basic engineering solidity don’t happen overnight.

There is now an ecosystem of companies building tooling to support Docker in production- startups like ClusterHQ, Datadog, Rancher Labs, Server Density, Sysdig, Treasure Data and Weaveworks. More established players such as AppDynamics, CloudBees and New Relic. Also of course the cast of existing suppliers looking to embrace and extend Docker- including Amazon Web Services, IBM, Microsoft, Oracle, Pivotal etc. Then there are outright competitive plays for the bigger prizes, such as Google Kubernetes. There will be negative commentary – growing pains are par for the course.

Docker is not going to have everything it’s own way – but the path is now set clear for Docker to become an industry standard production platform. We can call the path it’s on the Docker Pattern.

For further reference see also
IBM, Red Hat adopt “VMware Pattern” for Cloud. Disruption Strategy Emerges
Amazon Web Services: an instance of weakness as strength

Our clients include Docker itself, Amazon Web Services, IBM, New Relic, Oracle, Pivotal and Treasure Data.

Debbie Clapper made the beautiful pattern above.

15 comments

  1. I frequently find myself explaining why Docker is being adopted (in the Enterprise) so much quicker than VMs. It seems to be that VMs got tied to hardware upgrade cycles (typically 3 years) whilst containers haven’t got caught up in that. So what took 10yrs with VMs looks like it will take 6yrs with containers (and we’re already 2yrs in – the early adopters are done already and we’re moving into the majority).

  2. >I frequently find myself explaining why Docker is being adopted (in the Enterprise) so much quicker than VMs.

    Is it though? If you talk about VMs in the sense of enterprise server x86 virtualization (ESX–basically), that came out in 2002 and virtualization was pretty widespread within the next 5 years or so.

    There’s certainly a huge amount of interest in Docker–I agree with James there–but I’m not sure I expect to see production deployment happen a lot faster than hardware virtualization.

    In fact, I’d argue that it wouldn’t be surprising to see adoption happen a bit slower–whatever the ultimate impact–given that containerization causes more things to change than VMs did.

  3. I find it a little funny that you cited my uber-positive article on Docker adoption as “negative commentary,” James. 🙂 Only the headline was negative, but I didn’t write that.

  4. @mjasay you know most people only read the headline, matt. 😉

  5. […] have never seen a technology become ubiquitous so quickly [as Docker],” RedMonk analyst James Governor declares. Red Hat CEO Jim Whitehurst has also joined the Docker fray, telling investors that Docker is the […]

  6. […] have never seen a technology become ubiquitous so quickly [as Docker],” RedMonk analyst James Governor declares. Red Hat CEO Jim Whitehurst has also joined the Docker fray, telling investors that Docker is the […]

  7. […] have never seen a technology become ubiquitous so quickly [as Docker],” RedMonk analyst James Governor declares. Red Hat CEO Jim Whitehurst has also joined the Docker fray, telling investors that Docker is the […]

  8. […] Rise of the Docker Pattern, RedMonk – Cogent breakdown of Docker’s significance, comment in comments about hardware lifecycles is legit. […]

  9. […] monitoring container infrastructure for quite some time. James Governor has discussed the “Rise of the Docker Pattern,” and New Relic is in a great position to assess this pattern’s […]

  10. Thanks… very well written article.

    “Amazon is the new VMware”…… but adopting faster is the crux.

    The displacement of VM’s with containers only plays out if “bare metal” for containers proves common place & I don’t think that will happen in the Enterprise. Personally I think containers running in VM’s are the better way to go for most. Solves networking probs (not just NAT to host, but NSX can IP each container for NFV use cases & security) and monitoring tooling without new silos… not to mention 99% of folks will have type-2 & 3 apps for a long time… why introduce all this mess of new tooling & build a dedicated infrastructure stack.. be it on prem or in public cloud when you can have one integrated one?

    Moreover, when you think about it… with reliable persistent storage in containers now you can take more Type-2 apps and containerize them because a container running in a ESX VM is no more or less resilient than just running 1 app per VM but now you don’t have all those duplicate O/S’s that take up resources & need to be patched all the time…. And of course all the Docker development benefits.

    Human nature 😉 which one will win… when sometimes its about how they will weave together.

  11. […] building with an expectation of disposability. With the Docker Pattern of course you’re going to throw stuff away. Enterprises are now getting the message about minimum […]

  12. […] (again, relative) ease of use has spurred enterprise adoption, according to a separate essay by RedMonk: “Unlike virtual machines, containers include the application and all of its dependencies, but […]

  13. […] (again, relative) ease of use has spurred enterprise adoption, according to a separate essay by RedMonk: “Unlike virtual machines, containers include the application and all of its dependencies, but […]

  14. […] RHEL for IBM Z using the s390x architecture. So now it’s Docker’s turn. We’ll likely see the Docker Pattern assert itself at companies like ADP (big Docker shop, big mainframe shop) and Walmart (mainframes, […]

Leave a Reply to Gordon Haff Cancel reply

Your email address will not be published. Required fields are marked *