“In God we trust, all others must bring data” ~ W.E. Deming
The rise of docker as a tool to package & run software has been nothing short of an urban legend coming true. In this article I’ll use data points to showcase the sheer magnitude of the success of the technology & highlight the trends of container adoption across enterprises.
This article is the first part in a series of posts, in the next article i’ll discuss the challenges that organizations face while trying to adopt containers.
Docker is an open platform to build, ship, and run distributed applications on laptops, data center VMs, or the cloud. It makes it possible to separate the applications from the infrastructure and also to package & run an application in a loosely isolated environment called a container. These attributes enable teams to deliver software faster & more efficiently.
The best way to understand the growth of the docker platform is to look at Figure’s 1.0 & 1.1. These charts provide a direct indication of the users interacting with the docker platform, by plotting the cumulative sum of images downloaded from the official docker store over time.
As is evident from Fig 1.0, it took about two years (somewhere in the mid of 2015) for the technology to reach the Tipping Point, post which the adoption has grown in an exponential manner.
At the time of writing this article about 1 Billion images were being downloaded from the online docker store every 2 weeks & around 38 Billion images have been downloaded till date. In the last one year more than 1 Million new developers jumped onto the docker bandwagon.
Adoption rates such as these are almost unprecedented in the software industry & were partly fueled by the founder’s strategy to open source the technology early in its life cycle. Today a large active community is contributing to the success of this platform
Fig 1.1 (source: blog.docker.com)
We would have to go way back to the beginning of the millennium to find another software that met with such super success early in it existence. It was in early 2000 that Salesforce flew out of the blocks after the maverick & controversial “End Of Software” advertising campaign and has never looked back since.
The great thing about docker is that it alleviates pain points for technologists across all spectrums of the SDLC. As a result Docker found early promoters in the best of breed architects, developers, testers as well as operations engineers.
These early adopters were keen on sharing the knowledge they had gained and desired to collaborate with like minded individuals to enable the success of the technology. Hackathons were organized, meet-ups were planned, blogs written, online webinars created. In short everything possible was done to share the knowledge! The wider community came on board and used the technology to resolve their existing pain points, experienced the benefits and pushed for adoption within their organizations.
What followed was a classic Pull based technology adoption across organizations of all sizes and businesses, as depicted below in Fig 2.0.
Fig 2.0
As the the container technology became exceedingly popular the eco-system around it also began to grow at an astounding pace. “The force was so strong” with docker that all the technology behemoths & bellwethers of the day jumped into the fray and started investing in creating their own platforms for providing docker based services, products or managed platforms.
Fig 2.1 (source: blog.docker.com)
By now most enterprises had already started experimenting with the technology in non-production environments in some form or the other. As the eco-system matured, more and more production workloads were being moved over to containers, traditional applications were being containerized & a very high percentage of new applications were designed to be docker compatible.
As of 2018, docker has proved its production capabilities and based on a report by Datadog around 25% of their enterprise customers have adopted Containers in production & only 8 % have abandoned the technology (Fig 3.0). Most of the organizations that are dabbling with the technology also become adopters in the long run. The adoption trend line has consistently stayed positive and shows no signs of tapering, at this rate around 45–50 % of all enterprises would be using containers in production by 2021.
Fig 3.0 (source: Datadog)
What is extremely important to note here is that once an Enterprise decides to go with containers, the footprint of the workloads in containers keeps on growing at a breakneck pace, with the deployment rate increasing by 75 % in the last one year alone as depicted in Fig 3.1. Also noteworthy is the fact that the deployment rate has been increasing year on year. This data point is in itself a testimony to the value that containers bring to an enterprise.
Fig 3.1 (source: Datadog)
While the use cases for containers can differ from enterprise to enterprise, the benefits accrued tend to be consistent if the container strategy is well thought through. Use of docker brings substantial improvements to MTTR (~ 60 %), Developer Productivity (~65%) & reduction in the no of VM’s & server costs (~ 45%). These improvements would lead to much faster time to market, better user experience & lower costs.
Fig 3.2 (source: state of App Development survey)
Another important aspect of the adoption of the containers is the fact that most Enterprises are using managed container platforms to deploy containers in production across on-Premise and cloud based setups (Fig 3.3). It can also be inferred from the below data that most enterprises are choosing to use the managed container platform of the cloud vendor that they are already engaged with. While Kubernetes & Docker Swarm have emerged as the dominant orchestrator engines for running production workloads.
Fig 3.2(source RightScale )
As the container adoption grow and the eco-system keeps on evolving it would become increasingly important to learn for the missteps and create blue prints for successfully moving to a containerized environment & that is what the focus of the existing IT teams should be.
In the next article I discuss the challenges faced by organizations while adoption containers.Find the link here Challenges in Container Adoption.