As software development continues to grow with the world around us and offers more stable and lucrative opportunities, new developers might wonder how to differentiate themselves in an industry that is fairly homogenous and stereotypical. How do you break the mold and stand out from your peers to find more success and happiness in your career?

I was recently working with a collegue to deploy a Sitecore app to Azure via Pipelines. The release was running successfully, but the deployed files didn’t seem to be overlaying the required, pre-loaded Sitecore files. When we viewed /site/wwwroot from Kudu, we were seeing only the files from the build artifacts; when we viewed the same directory from FTP, we were seeing only the Sitecore files. What gives?

Azure AKS with RBAC

Over the past months, I have been digging into Kubernetes pretty hard. Since most of my clients are using Azure, AKS is currently the managed Kubernetes service of choice in my world. Since it has gone GA in June and now supports RBAC, I have been looking to provision a Kubernetes playground for the broader organization.

So I have this React application bundled with webpack and containerized in an image based on an Apache httpd base image. I want the app to make synchronous HTTP requests, but I don’t want the container at build time to be aware of where it’s making that request - this should be a configuration detail rather than implementation.

In a microservice architecture, it’s common to split an API into multiple, independently-deployable applications. While technically the services that run in front of the application containers can expose themselves directly to the internet via ports, it’s best practice to serve traffic traffic from a cluster over ports 80/443 and route internally via DNS.

Size matters. When you’re shipping code in containers, it’s important to remember that your container images are pulled on every node within your Swarm or Kubernetes clusters. This means that as you scale out your cluster nodes or schedule workloads on new nodes in your cluster, the size of your image becomes non-trivial for cold start time.

It’s a best practice to use containers to resolve dependencies (dotnet restore, npm install) and build (dotnet build, webpack) code because it ensures a consistent environment where the team’s specific version of Node.js becomes less important. Unfortunately, this also requires that the containers in which we build code must also have build dependencies installed, such as the .NET Core SDK or Node.js. Since you don’t need all those dependencies in production, how can you slim down your images?

The rate of change in software development tools has never been faster, which has allowed developers to continue to demand increased flexibility in how they deliver. Gone are the days of an entire development team sharing consistent hardware (PC vs Mac), and I have seen a massive uptick in once-Windows-only devs moving to MacOS. Enabling this change are the framework giants like Microsoft, who have been moving toward OS-agnostic frameworks - for example, .NET Core running on Windows, MacOS and Linux.

As technology companies continue to grow and project teams expand to deliver larger projects, it is important to reflect and make sure your organization’s foundation skills are strong. By no means would I say that learning Git will solve all of your problems, but I think it is fair to say that there are a few basic guidelines out there that keep your projects hovering at least closer to the “pit of success.”

These are guidelines that serve to document basic process around how I deliver all technical phases of our projects. While this process serves as the default, there are potentially reasons to change based on specific needs. If you do not follow this process, please be prepared to defend your decision(s) with your leadership.

The .NET world is hot on .NET Core right now, and why not? As the next iteration of the framework, it also allows developers more freedom to use their device and hosting platform of choice. While I use a PC at work, I have a Mac at home and love the idea of collaborating on projects from either device. .NET Core is the natural, easy way to do that.

The more time I spend in the tech community and market these days, the more it has become clear that the appetite for newer, more cutting-edge technologies has gone up. Lately, the new kids on the block that we’re tending to build more apps in right now are .NET Core 2.0 and Angular 4, and since we typically build internal applications, I’m seeing Azure AD used a lot for authentication. Since I am seeing this type of application so often, I thought it would be a good idea to build a reference architecture to tie it all together as a quickstart.

For reasons I’ll detail in a later post, I prefer to host my SPA and API applications separately. Combine that with the fact that I prefer to stay on the Microsoft stack and it raises the question: what is the best hosting environment for my Angular (or any SPA) application?

If you’re from the JavaScript/Node side of town, the answer is easy: Node is a great because it’s lightweight and configuration is very simple. Thankfully these days Microsoft has a real competitor in .NET Core – you get all the benefits of a serious application without the bloat that comes with a ASP.NET 4.x application. So now if I’ve bought in and decided on .NET Core, how do I actually serve my application? Let’s do a little bit of configuration to use the /wwwroot directory that comes with .NET Core applications.

I’m doing some modernizing of legacy apps and decided to use it as an opportunity to play with some new technologies - in this case, we’re talking about .NET Core 2.0 Preview 1. While building a reference architecture to talk about on Github, I ran into a snag while trying to figure out how to hide my Azure AD credentials. I want to include a config section, but don’t want to commit it with the rest of my appsettings.json file.

A few months ago, I posted about creating a decoupled, scalable web store order fulfillment app using an Azure Web App (as a webhook), WebJobs and storage queues. This is great for ensuring that orders don’t overwhelm your infrastructure, but it makes an assumption that queue consumers will handle the peak traffic over time and a single instance of these WebJobs is enough to handle all of your orders.

But what about when that’s not enough? In enterprise solutions with enough throughput, we need a way to keep up with demand. In this post, I will talk about the basics of Azure Functions and will follow up with a subsequent post that talks more about development, source control and deployment options.

Modern technology as we know it has become analogous to automation, and its distance from the consumer is decreasing. What was once reserved for production lines and the like has quickly expanded to handle everyday transactional workloads. I recently had such a challenge from a client wanting to decrease the time and effort to notify their warehouse of completed orders on their Shopify eCommerce platform.

As a developer, my favorite part of my job is that I get (and am encouraged) to play around with all kinds of new technologies to test for instances where we could add a new tool to Rightpoint’s ever-expanding toolbox. One of my latest finds is called Neo4j, an industry leader in the Graph Database space. While technically the product has been around for nine years, it seems that they’re just starting to become broadly relevant within the mainstream technology market.

It has been a while since I started tinkering with Neo4j, and today I got the opportunity to attend a Neo4j-sponsored Graph Database conference in Chicago. I would like to share a few learnings for those who want a TL;D[Attend] of the action. Firstly, let’s dive into a bit of quick context.

Introduction

While I was completing a project for a client of mine, I was working with a ListView that we wanted to create interactions on the screen, such as displaying details of the selected item in bound controls. We were using Prism, however, which naturally (and very deliberately) blocks the developer from making concrete references to Views from ViewModels. The issue that creates is that you cannot subscribe to events generated from a control like a ListView, which is exactly what we needed to do.

Introduction

In the last post in my Windows Universal App Series, I talked about building a Line of Business app for an amazing client of ours that is used for on-site inventory management with their respective client(s). One of the major challenges that our client faced was that when they went on-site, it’s typically in large buildings whose structures aren’t conducive to 3G/4G signal, and therefore wanted an application that would allow their reps to download data to their devices before beginning their work – while they are outside and had signal. They wanted the device to queue all changes and synchronize at the rep’s request once finishing their service call.

Introduction

I recently had the pleasure of working with a client who wanted to upgrade the platform of their Windows Mobile 6.5 Line of Business, inventory management app to something a bit more recent and cutting-edge. Their goal was to modernize their devices (the old device was a Motorola ES400), and update their software to something significantly more maintainable, scalable and extensible.