Azure Services for Microservice Developers

The Microsoft Azure landscape can be huge and overwhelming if you don't know where to start. In this session I'd like to focus on a few services that can be of additional value to your new or existing microservices. Whether you need an api or a worker, Azure is your bet.

When you move your code to the cloud in an online repository you don't want hardcoded credentials in your codebase. Key Vault is a password safe that integrates fully with the Microsoft stack.

There are different services for messaging (events & queues) that come native in the Azure stack. Event grid and Service Bus are easy to start with and powerful in features.

If your microservice is tiny enough you might want to consider a nosql database. Mssql can be overkill if you just need one table. Azure Cache for Redis and Azure Cosmos DB can be interesting alternatives.

Finally, you can control exposure of your API with API Management and I'll briefly touch on Application Insights for monitoring your deployment.

Knowing your options is important when architecting your next application. This overview will help you navigate through the flood of Azure services that are around.

My Experiences with Monoliths and Microservices

In the past years I've collaborated to different microservice architectures. I'd like to share the things I did and learned. Monoliths are not bad by default and equivalently microservices are not good by default. They both have their pros and cons. Each software architecture will need its own design. There is no one best fit for all.

You can get yourself into microservices in two ways: either you break a monolith into microservices or you start from scratch and create a microservice architecture.

Breaking a monolith means locating the inner dependencies. I'll show some strategies on how to decouple a monolith and where to start cutting. Starting from scratch means you'll need great knowledge about the future usage of your product.

I have taken both paths to microservices and I'd like to talk about my own experiences along the way.

Google and Amazon cloud for the Azure developer

Going to the cloud is a given. Endless possibilities with over 200 managed services, global coverage with datacenters everywhere and you are charged per usage. But which cloud should you use? As a Microsoft developer, you'd choose Azure right?!
In this session I'll show you the similarities between the three major clouds Microsoft Azure, Amazon AWS and Google GCP. I'll argue that in the end there isn't too much difference. The only thing that matters is that you should see the whole picture.

Help! They want me to use Linux!

In the recent years, Microsoft has become more aware of Linux and open source software. .Net Core is a big hit, Sql Server runs on Linux, they open source everything. You’re using or deploying other people’s software and now you’re asked about Linux. But how are you as a Microsoft minded developer fitting in to this all. You’re quite comfortable in your all Microsoft setup and now you need to use Linux. HELP!

In this session, we'll use SSH to connect to a linux machine; show the basic commands; deploy/copy some executable or binaries to the machine; and finally, run an executable using Mono.Net and .Net Core.
You'll learn that Linux is easy and there is nothing to be afraid of.

duration: 45 min; audience: Microsoft .net developers; first public delivery: .NET Zuid (Netherlands)

Simplify running your VM applications by using Docker in Azure Container Instances

Want to use Docker, but having trouble on where to start? Welcome Azure Container Instances! The ‘serverless’ VM that really amazed me in its completeness. I’ll show you how to move your legacy VM setup to Azure Container Instances, so you can start rocking with Docker!

The starting point will be a Windows VM running somewhere with your .NET software on it. Maintenance of the VM (whether on premise or cloud) lies in your hands and deploying new versions of your tool is difficult to automate. I’ll show you how to run your setup in Docker on your local machine and how to ship the entire Docker container with Azure Container Registry. Once shipped, you can easily deploy the Docker container with your app to an Azure Container Instance. Finally, I’ll walk you through the Azure Container Instances features in the Azure Portal: you can see logging, inspect Environment Variables and start a CLI session from within your browser.

first public delivery: SDN Zeist (NL) June 2019

OK, I’ve created something, now you make it scalable and run in Azure

The context: you have both C# and C++ applications running on respectively Windows and Linux operating systems. The goal: integrate all this into a pipeline where the one is generating input for the other. You’ll need a layer to facilitate this.

You will learn how you can create a processing flow that is cross platform, scalable and cloud ready.

We will walk through an approach where the applications are strictly separated from the flow management. Nodes will have a single task with the smallest of jobs to aim for maximum scalability. Furthermore, nodes will have a minimum of dependencies to be able to deploy it everywhere. In the end, everything is a black box.