What kind of submission are we looking for in the Architecture track at Devoxx UK? With less than a week left to submit your proposal, if you’re struggling for inspiration, this is the article for you. With thanks to Program Team members Gayathri Thiyagarajan, Kate Stanley and Luke Whiting.
What do you want to learn more about?
Complex Serverless Systems
Luke: We saw a lot of intro to serverless type talks in the last couple of years. I think most of the industry is comfortable with the concept now and knows how to deploy a trivial function and attach it to a web endpoint. What I would be interested in seeing is how you build a complex system with serverless. How do you manage all those individual functions, how do you deploy them, how do you interact with databases ETC? It would be great to see some real war stories of people who have products or large chunks of products running in such environments.
Kate: Also I am interested to hear where people are going with serverless and event-driven. As those have been popular subjects recently. Specifically incorporating serverless or event-driven into existing monoliths or REST based microservices rather than building from scratch.
Microservices in the wild
Kate: It would be interesting to hear about how companies that are providing an installable product (rather than providing services) are architecting their product. Are they building in a microservice architecture but then releasing as a monolith? Are they still just building monoliths? I think a lot of talks focus on the “ideal” use case for microservices, rather than tackling the areas where a product can’t be delivered in microservices.
Gayathri: Of late the Microservices world seems to have gone into the depths of tech stack such as Service Mesh, Container Orchestration – I would like to know if there has been any new developments in architectural patterns around Microservices.
Non x86 programming
Luke: I would also like to see something on non x86 programming: dedicated AI chips (IE Googles Tensor Processing Unit), ARM based high core count servers, GPUs and FPGAs – these are all becoming more and more common.
Data Platform architecture
Gayathri: I think Data Platform architecture topics are a bit underrepresented. I have read a few blogs on how Netflix, Uber and AirBnB do this – but not much from anywhere else. I would also like to hear more on Machine Learning or Data Science platform architectures.
Reactive and WebFlux
Luke: Finally I’d like some more talks on WebFlux and reactive programming. I’m still not sure how well that architectural philosophy will take off so it would be nice to see more about it’s advantages and disadvantages.
What are your predictions for 2020?
Gayathri: Micro/modular frontends were quite the buzz in 2019, I think we will hear more on this subject, particularly war and success stories in 2020. Microservices are still squarely at the forefront – with more and more developers moving towards it – I expect some talks mainly around integration aspects if it.
Luke: I think the biggest shift we’re going to see is the continued move away from the traditional server-focused way of thinking about deploying apps towards platform centric deployment. Things like the announcement of Amazons EKS Fargate where you have a Kubernetes cluster with no worker nodes: instead all K8S services spin up their containers in a Serverless environment. This shows that a move away from thinking about machines to much higher levels of abstraction is becoming the norm.
Kate: I’m expecting to see a lot of submissions around micro-frontends as that seems to be a growing theme. Also talks around the realities of building Microservices now that we are firmly in a place where many companies have Microservices in production.
You have until 13 Jan 2020, 23:59 GMT to submit to the Call For Papers. Hurry!