It has been a while since I returned from DockerCon Europe in Copenhagen, and despite doing a presentation 4 days after arriving home for Docker Cambridge meetup group, I haven't yet done a blog post for it. However, that is about to change. So, here is my review of DockerCon Europe.
Overall, what I love to see in software is passion and ideas around everything. I remember feeling how exciting it was to be involved in a massive open-source project like Drupal in 2011 at DrupalCon London and to me it seems that Docker is around the same stage of growth now (in terms of number of attendees), however, its potential as a software solution is far different to Drupal. My involvement with Docker started seriously around 2 years ago, when looking for scalable solutions for deploying the same parts of software and I am probably even more committed to it as a solution than I was in 2015. DockerCon left me with 3 things:
- The community is engaged and rapidly iterating
- The product (Docker/Containers etc) is gaining massive adoption
- More companies are now supporting the container ecosystem
Unless you have been living under a metaphorical rock, you will know the big news from DockerCon. That Kubernetes is going to have full support in Docker from next year. The tools for deploying to Kubernetes or Docker Swarm will be the same. This has the massive benefit of mitigating some of the complexity of managing Kubernetes, but also focussing toolsets back into Docker, which I guess is good for the company.
I think that the really interesting part of this decisions is that Docker has tried to run an orchestration platform (Swarm) and accepted that for some use cases people were preferring Kubernetes and they have rapidly evolved their product to accept that. In real terms, the engineers have listened to the wider community and brought in the most wanted tools.
Docker Gaining Adoption
A big thing for Docker is how it can going customers and users from an already active market. The benefits in resource use may be great, but when you are risking large international systems, with large and lengthy release and software pipelines it can be hard to get anyone to take a risk. Startups have far less technical debt and risk-aversion to just deploy containers. So, Docker is attempting to group the retooling of applications under MTA (Modernising Traditional Applications).
In many ways MTA and the program that surrounds it is a marketing exercise, to inject some more modern software practices into an industry which under estiamtes spends the vast amount of its budget on maintenance rather than innovation (between 60%-80% depending on who you ask is being spent on maintaining existing applications rather than building new ones). However, in a joint marketing effort, getting large players such as IBM, Microsoft, Accenture, Cisco to partner with them for deploying technology using containers is a big win.
The Container Ecosystem
The container ecosystem (which I am judging Docker to be a large player in) is a rapidly changing space. Not only have Azure now got 4 different products that use this (ACI, AKS, ACS and Azure Web Apps for Containers), but there is also a growth in systems that operate in just one area of container ecosystem, which I would judge to include; security, monitoring, storage, pipelines, developer tools, hosting. I am certain that I am not the only one who views the container ecosystem as a new paradigm in software development in our search for better resource usage and scale. But having new companies and existing companies really growing their offering for what is available to container users not only supports teams to professionalise their software stack, but also adds more opportunity to innovate on particular areas.
A Special Mention Of Other Areas
I would also like to just have a mention of some other really good parts of the conference. Firstly, as I am a community leader (because I run the Docker Cambridge Meetup) it was brilliant to network and get guidance from similar people doing a similar role from round the world (Melbourne, Manchester and Mannheim!). Also, I am particularly interested in the OpenFaas project which appears to be blending probably the hottest two ideas in software at the moment; Containers and Serverless.