DEV Community: Shayne Boyer

🚀 Build Your First Remote MCP Server on Azure Using TypeScript — and Supercharge Your AI Workflows

Shayne Boyer — Wed, 26 Nov 2025 17:30:56 +0000

If you're experimenting with the Model Context Protocol (MCP), you've already seen how powerful it is for connecting tools, agents, and applications into unified workflows.

But the real magic happens when your MCP server is not running on your laptop… but deployed in the cloud.

Azure now provides first-class support for Remote MCP Servers built in TypeScript and JavaScript, giving you scalable, secure, team-ready capabilities for AI-integrated workflows—from GitHub Copilot to custom agent frameworks.

This post covers why Remote MCP Servers matter, how Azure Functions make them simple, and where to start with official docs and sample repos.

🌩️ Why Build a Remote MCP Server?

Local MCP servers are great for quick prototyping, but teams hit real limits:

Local servers can’t scale
They aren’t shareable
They break inside containers and DevBoxes
They can’t reach VNet-secured resources
They rely on local credentials
Long-running tasks are fragile

Remote MCP Servers on Azure solve all of this.

Running in Azure Functions gives you:

Serverless scale
Enterprise-grade security
Managed Identity (no secrets on laptops)
Team-wide reusability
Integration with Azure services (Key Vault, SQL, Cosmos, Storage, Container Apps, etc.)

If you're building anything “agentic,” this unlocks a new world of capabilities.

📘 Start With the Official Microsoft Docs

Azure has three key resources to help you get started:

👉 Full Tutorial (TypeScript)

Build a Remote MCP server on Azure Functions
https://learn.microsoft.com/en-us/azure/azure-functions/functions-mcp-tutorial?tabs=mcp-extension&pivots=programming-language-typescript

This covers:

Setting up your project
Writing your first MCP action
Deploying with azd
Securing access
Testing from VS Code

👉 Quickstart: Custom Remote MCP Server

https://learn.microsoft.com/en-us/azure/azure-functions/scenario-custom-remote-mcp-server?tabs=linux&pivots=programming-language-typescript

Perfect if you already understand Functions and want the shortest path to success.

👉 Official Sample Repo

https://github.com/Azure-Samples/remote-mcp-functions

Includes:

Multiple MCP endpoints
Schemas & patterns
Deployment templates (azd)
Logging + error-handling examples
A recommended project structure

Use this repo as your starting point when building your own server.

🛠 How It Works (Simplified)

A Remote MCP server is just an Azure Function that returns MCP-compliant JSON.

Here’s a minimal TypeScript example:

// Hello function - responds with hello message
export async function mcpToolHello(_toolArguments:unknown, context: InvocationContext): Promise<string> {
    console.log(_toolArguments);
    // Get name from the tool arguments
    const mcptoolargs = context.triggerMetadata.mcptoolargs as {
        name?: string;
    };
    const name = mcptoolargs?.name;

    console.info(`Hello ${name}, I am MCP Tool!`);
    return `Hello ${name || 'World'}, I am MCP Tool!`;
}

// Register the hello tool
app.mcpTool('hello', {
    toolName: 'hello',
    description: 'Simple hello world MCP Tool that responses with a hello message.',
    toolProperties:{
        name: arg.string().describe('Required property to identify the caller.').optional()
    },
    handler: mcpToolHello
});

Deploy to Azure:

azd up

Now your MCP server is reachable from:

VS Code MCP Extension
GitHub Copilot
LangChain agents
Custom MCP clients
Automation pipelines and CI/CD

💡 What Developers Are Building Today

🔹 Cloud & infra automation

Validate Bicep, deploy Container Apps, check resource health.

🔹 DevOps workflows

Run tests, inspect containers, automate PR tasks.

🔹 Secure enterprise data connectors

Provide scoped, read-only access to internal SQL, Storage, Key Vault, etc.

🔹 Knowledge tools

Chunk, embed, and search documents stored in Azure.

🔹 Multi-step orchestration

Expose “one-call” MCP actions that run entire workflows under the hood.

🌐 Why TypeScript + Azure Functions Is a Perfect Match

Type-safe MCP schemas
Serverless autoscale
Zero infrastructure
Managed Identity → no secrets in your MCP client
One-command deploy (azd up)
Great for local dev, Codespaces, or remote teams

If you're already writing Node/TS, this feels natural and productive.

🧪 Try It Yourself — And Give Feedback

This is an evolving area—and Microsoft is actively looking for real developer feedback.

If you're building:

custom AI tools
cloud automation
team-wide secure connectors
multi-tool agent solutions

Now is the perfect time to explore Remote MCP Servers and help shape the next wave of agentic development.

🙌 Wrap-Up

Remote MCP Servers on Azure give developers cloud-grade scale, identity, and reliability, and TypeScript support makes them incredibly easy to adopt.

Start building now:

If you make something cool—publish it!
If you hit a rough edge—open an issue.
If you have ideas—share them loudly.

We’re still early in this space, and your creativity helps shape the direction of MCP and cloud-based agentic workflows.

Building Container Images Without Docker: Introducing pycontainer-build

Shayne Boyer — Mon, 24 Nov 2025 18:36:28 +0000

What if you could build production-ready container images for your Python projects without installing Docker, writing Dockerfiles, or dealing with daemon dependencies? That's the vision behind pycontainer-build — a native, Docker-free container image builder for Python.

The Problem with Traditional Container Builds

Today, containerizing Python applications typically requires:

Installing Docker Desktop or Docker Engine — Not always possible in locked-down corporate environments, cloud IDEs like GitHub Codespaces, or CI/CD runners
Writing and maintaining Dockerfiles — Boilerplate, multi-stage builds, and keeping base images updated
Understanding Docker-specific concepts — Layers, build contexts, caching strategies
Docker-in-Docker in CI — Complex and fragile setups with privileged containers

These friction points slow down developer onboarding and create unnecessary complexity for a straightforward task: packaging Python code into a container.

Enter pycontainer-build

Inspired by .NET's native container support and tools like Jib (Java) and ko (Go), pycontainer-build provides a Python-native way to create OCI-compliant container images.

Here's what it looks like:

# Install
pip install -e .

# Build a container image
pycontainer build --tag myapp:latest

# That's it!

No Dockerfile. No Docker daemon. Just pure Python creating OCI images.

How It Works

pycontainer-build introspects your Python project and automatically:

Detects your Python version from requires-python in pyproject.toml
Selects the right base image (e.g., python:3.11-slim)
Finds your entry points from [project.scripts] in pyproject.toml
Packages your application files into layers
Creates an OCI-compliant image layout ready for registries

Under the hood, it:

Uses only Python stdlib (no external dependencies!)
Creates proper OCI image manifests and configs
Generates content-addressable layers (SHA256 digests)
Supports pushing to any OCI registry (GHCR, ACR, Docker Hub, private registries)

Real-World Example

Let's say you have a FastAPI application:

# app.py
from fastapi import FastAPI

app = FastAPI()

@app.get("/")
def read_root():
    return {"message": "Hello from pycontainer!"}

# pyproject.toml
[project]
name = "myapi"
version = "0.1.0"
requires-python = ">=3.11"
dependencies = ["fastapi", "uvicorn[standard]"]

[project.scripts]
myapi = "uvicorn app:app --host 0.0.0.0 --port 8080"

Build and push with one command:

pycontainer build \
  --tag ghcr.io/myorg/myapi:latest \
  --include-deps \
  --push

pycontainer-build will:

Pull python:3.11-slim as the base image
Package your app.py and dependencies
Configure the entrypoint to run uvicorn
Push the image to GitHub Container Registry

Compare this to a typical Dockerfile approach:

# Dockerfile
FROM python:3.11-slim

WORKDIR /app

# Install dependencies
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

# Copy application code
COPY . .

# Expose port
EXPOSE 8080

# Run the application
CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8080"]

docker build -t ghcr.io/myorg/myapi:latest .
docker push ghcr.io/myorg/myapi:latest

The Dockerfile approach requires Docker installed, manual maintenance of the Dockerfile, and explicit knowledge of best practices (multi-stage builds, layer optimization, etc.). pycontainer-build handles all of this automatically.

Key Features

🎯 Zero Docker Dependencies

Pure Python implementation using only stdlib. Works in environments where Docker isn't available or allowed.

🧠 Smart Auto-Detection

Python version from pyproject.toml
Entry points from [project.scripts]
Project structure (src/, app/, or package directories)
Web frameworks (FastAPI, Flask, Django)

📦 Base Image Support

Build on top of official Python images, distroless, or custom base images. Properly merges layers and configuration.

🚀 Registry Integration

Push to any OCI-compatible registry with automatic authentication:

GitHub Container Registry (GHCR)
Azure Container Registry (ACR)
Docker Hub
Private registries

⚡ Layer Caching

Content-addressable caching with LRU eviction. Only rebuild layers that changed.

🔒 Security & Compliance

SBOM generation (SPDX 2.3 or CycloneDX 1.4)
Reproducible builds with deterministic timestamps
No secrets in images (proper credential handling)

🔌 Toolchain Integrations

Poetry plugin: poetry build-container
Hatch plugin: hatch build
GitHub Actions: Reusable workflow
Azure Developer CLI: Custom build hooks
VS Code extension: Build from command palette

Use Cases

1. CI/CD Without Docker

# GitHub Actions
jobs:
  build:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
      - run: pip install pycontainer-build
      - run: pycontainer build --tag ghcr.io/${{ github.repository }}:latest --push

No need for Docker daemon or privileged containers.

2. Local Development

# Quick iteration
pycontainer build --tag myapp:dev
podman run -p 8080:8080 myapp:dev

3. Multi-Service Applications

# Build multiple services
pycontainer build --context ./api --tag myapp-api:latest
pycontainer build --context ./worker --tag myapp-worker:latest
pycontainer build --context ./frontend --tag myapp-frontend:latest

4. Restricted Environments

GitHub Codespaces (no Docker daemon)
Azure Dev Box
Corporate locked-down workstations
Air-gapped environments (offline builds)

Configuration Options

Command-Line Interface

pycontainer build \
  --tag myapp:v1.2.3 \
  --base-image python:3.11-slim \
  --include-deps \
  --workdir /app \
  --env KEY=value \
  --platform linux/amd64 \
  --sbom spdx \
  --config pycontainer.toml \
  --verbose \
  --push

Configuration File (`pycontainer.toml`)

[build]
base_image = "python:3.11-slim"
workdir = "/app"
include_deps = true
reproducible = true

[build.labels]
maintainer = "team@example.com"
version = "1.0.0"

[build.env]
PORT = "8080"
ENV = "production"

[registry]
url = "ghcr.io/myorg/myapp"

Python API

from pycontainer.config import BuildConfig
from pycontainer.builder import ImageBuilder

config = BuildConfig(
    tag="myapp:latest",
    context_path=".",
    base_image="python:3.11-slim",
    include_deps=True,
    env={"ENV": "production"},
    generate_sbom="spdx",
    reproducible=True
)

builder = ImageBuilder(config)
builder.build()

Perfect for integration with build tools, AI agents, and automation scripts.

Under the Hood: OCI Image Spec

pycontainer-build creates compliant OCI images following the OCI Image Specification:

dist/image/
  ├── index.json              # OCI index (manifest list)
  ├── oci-layout              # Version marker
  ├── blobs/sha256/
  │   ├── <manifest-digest>   # Manifest blob
  │   ├── <config-digest>     # Config blob
  │   └── <layer-digest>      # Application layer (tar)
  └── refs/tags/
      └── <tag-name>          # Tag reference

Each layer is:

Created as a tar archive with proper paths (/app/src/, /app/pyproject.toml)
Compressed (optional)
Hashed with SHA256 for content-addressable storage
Referenced in the manifest by digest

This means images are:

Portable — Work with any OCI-compatible runtime (Docker, Podman, containerd, CRI-O)
Cacheable — Layers shared across images save space
Verifiable — SHA256 digests ensure integrity

Comparison with Other Tools

Feature	pycontainer	Docker/Dockerfile	Jib (Java)	ko (Go)
No Docker daemon	✅	❌	✅	✅
Language-native	✅ Python	❌ Generic	✅ Java	✅ Go
Zero config	✅	❌ Requires Dockerfile	✅	✅
Auto-detects dependencies	✅	❌	✅	✅
Pure stdlib	✅	N/A	❌	✅
Python-specific optimizations	✅	❌	N/A	N/A

Current Limitations

This is an experimental project. Known limitations:

Multi-arch builds: Platform flag sets metadata only; no actual cross-compilation yet
Framework detection: Supports FastAPI, Flask, Django (easy to extend)
SBOM scope: Python packages only (doesn't parse OS packages from base images)

These are planned for future phases. Contributions welcome!

Getting Started

Installation

# Install from source (PyPI package coming soon)
git clone https://github.com/spboyer/pycontainer-build.git
cd pycontainer-build
pip install -e .

Quick Example

# Create a sample app
mkdir myapp && cd myapp
echo 'print("Hello from container!")' > app.py
cat > pyproject.toml << EOF
[project]
name = "myapp"
version = "0.1.0"
requires-python = ">=3.11"
EOF

# Build container
pycontainer build --tag myapp:latest

# Output location
ls dist/image/

Push to GitHub Container Registry

# Authenticate (uses GitHub CLI or GITHUB_TOKEN env var)
gh auth login

# Build and push
pycontainer build \
  --tag ghcr.io/yourusername/myapp:latest \
  --push

Why This Matters

For Python Developers

Simpler workflow than writing Dockerfiles
Faster onboarding to containerization
No Docker Desktop licensing concerns
Native Python experience (feels like pip or poetry)

For DevOps Teams

Eliminate Docker-in-Docker complexity in CI/CD
Faster builds with intelligent caching
Security compliance with SBOM generation
Reproducible builds with deterministic layers

For the Python Ecosystem

Modern, standards-based approach to containers
Foundation for tool integrations (Poetry, Hatch, Azure DevOps)
Opens possibilities for Python in cloud-native environments
Aligns with .NET/Java container strategies (unified multi-language story)

Roadmap

Completed (Phases 0-4):

✅ Core OCI image generation
✅ Registry push support
✅ Base image support
✅ Layer caching
✅ SBOM generation
✅ Toolchain integrations (Poetry, Hatch, GitHub Actions, VS Code)
✅ Framework auto-detection

Coming Soon:

🔜 Full multi-arch builds (ARM64, AMD64)
🔜 PyPI package distribution
🔜 More framework support (Streamlit, Gradio, etc.)
🔜 Build optimization (parallel layer creation)
🔜 Integration with more CI/CD platforms

Try It Out and Share Feedback!

This project is experimental and evolving based on community feedback. I'd love to hear:

What use cases resonate with you?
What blockers do you face with current Docker workflows?
What features would make this production-ready for your team?

Links:

📦 Repository: github.com/spboyer/pycontainer-build
💬 Feedback & Issues: github.com/spboyer/pycontainer-build/issues
📖 Documentation: See README for more details

If you find this useful, give it a star ⭐ on GitHub and share your experience!

Acknowledgments

Inspired by:

.NET SDK's native container support
Jib (Java) — Daemonless container builds
ko (Go) — Simple container images for Go

Unveiling VS Code for Web - Azure

Shayne Boyer — Mon, 11 Nov 2024 18:29:54 +0000

Watch the live stream replay - View the VS Code You Tube

Welcome to the freshly minted VS Code for Web - Azure, a development environment designed to simplify your Azure app building experience. This synergistic blend of VS Code and Azure is geared towards developers, offering a universal environment powered by Azure, to create, debug, and deploy applications. Let's dive into this transformative upgrade, its offerings, and how it revolutionizes the development experience.

Why Choose VS Code for Web - Azure?

VS Code for Web - Azure is a potent tool for developers looking to deploy their apps using Azure services seamlessly. Boasting a host of Azure's incredible features, this fully-fledged development environment is baked into the browser, unleashing the power of Azure development minus the intimidating setup process. The best part? It syncs seamlessly with your Azure subscription, bringing the universe of Azure services right at your doorstep.

Check out the discussion during an interactive live stream with insightful commentary by Olivia, a Developer Advocate on the VS Code team, Alex, an engineer on the Azure extensions team, and Meera, a Product Manager in the Azure Dev Tools team.

Watch the live stream replay - View the VS Code You Tube

One-click Provision and Deploy

Through the arsenal of curated Azure Developer CLI (azd) templates, you can effortlessly deploy a containerized app, configure CI/CD, or even launch your machine learning model. This simple Flask app or an intricate AI app is only a click away!

Check out the new AI Azure App Templates

Clicking on 'try template,' brings you to VS Code for the Web. You will be required to sign in to your Azure account, a one-time process that carries over to all your future sessions. With a swift click on 'one-click provision and deploy,' and choosing your Azure subscription and Azure location, you set in motion a series of automated processes, leading up to deploying the app. What's more, while your deployment is underway, you could get busy beneath the surface with the host of features enlisted in the walkthrough.

The Power of Azure on Your Browser

From conversing with GitHub copilot, a preloaded option, and an enviable service that suggests code and auto-completes your lines to managing your Azure resources, you are thrust into an all-in-one space for effective project management. Whether it's deploying your app using Azure CLI or using the Azure Developer CLI, the terminal employed in VS Code for Web - Azure is a wellspring of opportunities.

As a developer, you can now get a handle on all your Azure subscriptions, monitor and manage Azure applications, using CI/CD, build your app, and deploy it all within this single space! Notably, all these services are tethered to anchored compute, allowing for an array of tasks.

Saving Your Work

When dealing with a web-based coding environment, an immediate concern is retaining changes made to your code. Fortunately, you can save the changes to your projects in VS Code for Web - Azure by committing your code to GitHub. With just a few clicks, every line and letter could be saved to your GitHub repository.

Conclusion

VS Code for Web - Azure dispenses the need for bouncing between different applications, simplifies the initiation process, and mitigates the cognitive overhead associated with juggling various tools.

The private preview phase of VS Code for Web - Azure is an invite for feedback from developers worldwide. What's your take on this exciting new development environment, and what features would make your coding experience more enjoyable?

References

Azure Developer Tools
Azure Cloud Shell
GitHub Copilot
VS Code Azure Extensions

Streamline Your Azure Workflow with GitHub Copilot for Azure in VS Code

Shayne Boyer — Fri, 01 Nov 2024 18:08:03 +0000

I'm excited to share some great news with you all. We've just launched the public preview of GitHub Copilot for Azure - a fantastic new tool that seamlessly integrates with GitHub Copilot Chat in VS Code. Think of it as your personal assistant for navigating the Azure cloud.

See announcement post from Chris Harris!

Why GitHub Copilot for Azure?

Instead of constantly switching between your IDE and the Azure portal to manage infrastructure or search for commands and arguments, you can now focus on what you do best – writing code. Whether you're provisioning services or deploying apps, just ask @azure in GitHub Copilot Chat and handle it all right inside your editor.

Key Features

Learning about Azure: GitHub Copilot for Azure makes learning about Azure services a breeze. It pulls in relevant, up-to-date documentation like answers on Azure OpenAI models, Azure AI Search, or even how pricing works for services like Azure SQL, all within your coding environment. This is super handy for developers just starting with Azure, helping them grasp complex concepts faster, while also saving time for experienced developers who need quick reminders or details on the latest services and features.
Deploying Apps: GitHub Copilot for Azure takes the hassle out of deploying your apps by guiding you through tasks like setting up resources or automating deployments. Whether you're building a RAG (Retrieval-Augmented Generation) app with Python, creating a CI/CD pipeline, or using the Azure Developer CLI (azd) to deploy your project, @azure can suggest app templates, the right commands, and configurations straight from your code editor. No need to search for sample applications, look up CLI commands, or YAML syntax - @azure fills in the blanks for you.
Troubleshooting: Sometimes things go wrong. When they do, GitHub Copilot for Azure simplifies diagnosing and troubleshooting by providing quick insights into your application's performance and resource issues.

Get Started Now!

You can get started right now by installing GitHub Copilot for Azure from the VS Code Marketplace. It's designed to make your life easier and your workflow smoother.

.NET Learning Challenge!

Shayne Boyer — Sun, 01 Nov 2020 20:00:26 +0000

Compete, Learn, and Develop Skills

https://aka.ms/dotnetskills

November 1, 2020 - November 30, 2020

Compete against friends and coworkers by completing interactive modules on how to build apps across multiple platforms with .NET and earn your rank on the leaderboard.

Over 40 hours of FREE Learning inside of this awesome collection of 8 Paths and 5 additional bonus modules.

Topics include Basic C# skills, Web API development, Blazor Apps, Static Web Apps, Xamarin and Visual Studio or VS Code

Get started by going to https://aka.ms/dotnetskills

Learning Paths include:

Build .NET applications with C#
Add logic to your C# applications with C#
Work with data in C#
many more

Modules include:

Create a web API with ASP.NET Core
Improve the developer experience of an API with Swagger
Build a web app with Blazor Web Assembly

By the end of the challenge, you'll have marketable skills to better yourself and your career.

Tweet about your progress with the hastag #dotnetskills. Let me know @spboyer how you're doing. Challenge your friends and colleagues. How many modules and paths can you complete this month?

Have fun and learn.

ASP.NET Community StandUp : FAST Framework

Shayne Boyer — Mon, 13 Jul 2020 16:12:26 +0000

FAST is a collection of JavaScript packages centered around web standards, designed to help you efficiently tackle some of the most common challenges in website and application design and development.

In this episode, Rob Eisenberg sits with us to run through a number of the components, a Blazor web assembly example and answer questions about the what and why and talk about the roadmap for FAST!

Be sure to subscribe to the .NET Foundation Channel for more great .NET Content

Links from the show

Are you reading the "What's New" in .NET docs?

Shayne Boyer — Fri, 21 Feb 2020 14:47:32 +0000

Did you know that every month, the docs team at Microsoft has a post for what's new in the .NET documentation?

For each month there are new articles, updates to existing docs, and kudos for the community contributors.

It's super easy to get involved in contributing to the docs for the related technologies you are invested or interested in. See the Microsoft docs contributors guide or simply click the "Edit" link at the top of any doc to get started, they are all open source!

Here are some of the highlights forlast month .NET Docs

The Architecting Cloud-Native Apps for Azure architecture guide e-book added some new sections -

C# language revised the Value types and added new xml documentation

.NET added a great new article in the .NET Guide on How to migrate from Newtonsoft.Json to System.Text.Json

Other areas of the docs are also adding similar related what's new pages too. Check out ASP.NET Core docs updates and Xamarin updates too!

The docs continue to grow and evolve. Check in often on these pages to keep up to date with the latest content.

Updating the my blog from AppService to Containers...finally

Shayne Boyer — Fri, 17 Jan 2020 18:53:34 +0000

I have been running the well known ghost blogging platform on Azure AppService since ghost was version 0.10, about 8 years. In fact, it was before Azure had resource groups.

This means that it's Node, on Windows Servers and I am doing a bit of maintenance for upgrading the version of the platform. Running things like gulp.js scripts for builds and so and hoping for Kuduto do my git deployment after a couple of tries because I am running on a small box and there are limitations for networking and npm and start scripts to complete in enough time.

After a few successful upgrades, it became overwhelmingly painful to try and make the updates and life, work, etc. got in the way and now ghost is on version 3.2 and I am far to behind to try and do a manual update. So the jump to containers makes sense.

The data conversion was a little more work as I had to install ghost 1.0, import my data > export it and then import to 3.0.

Choosing the setup

I am using the latest ghost:3.2-alpine image available on Docker Hub.

$ docker run -d -p 3001:2368 ghost:3.2-alpine

It uses SQLite by default as the backend, but also offers MySQL which I'll use as an option through MySQL on Azure.

Analytics

I have been using Google Analytics for the past 12 years on my blog and want to continue with this solution. However, if you're starting a new solution, Application Insights is a quick click option when creating your AppService instance.

Create a new Dockerfile and add Application Insights in this manner:

FROM ghost:3.2-alpine

# Add app-insights globally
RUN npm install -g applicationinsights

# Link app insights to avoid rebuild/validate projects npm package tree
RUN cd current && npm link applicationinsights 

# ENV APPINSIGHTS_INSTRUMENTATIONKEY xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
RUN sed -i.bak -e "/var startTime = Date.now(),/a appInsights = require('applicationinsights'), " \
                -e "/ghost, express, common, urlService, parentApp;/a appInsights.setup().start();" current/index.js

You'll need to build your own Dockerfile and host it in a repository. i.e. Azure Container Registry, Docker Hub, GitHub etc.

This adds the npm package for Application Insights and adds the script to the index,js files of the application. Be sure to add your APPINSIGHTS_INSTRUMENTATIONKEY environment variable to your App Service Configuration.

Azure Web App for Containers

As I mentioned, the previous version was hosted in Azure App Service on Windows (Node 10 + Windows). Now we are moving to Linux (Alpine) on Web App for Containers.

Creating the app service was really simple. A single command line got the job started for me. (ok maybe 2, need a service plan too)

# Created my S1 service plan
az appservice plan create -n spboyerblog20 -g spboyer --is-linux - "East US" --sku S1 --number-of-workers 1

# Create the app service with the alpine 3.2 ghost image
az webapp create -n spboyer -g spboyer -p spboyerblog20 -i ghost:3.2-alpine

MySQL

I have setup the database backend using Azure Database for MySQL selecting bare minimums and an auto growth option - Basic, 1 vCore(s), 10 GB.

The data connections are set as ENV variables in my application settings.

Database settings for MySQL

Using a service for my data takes care of the potential data loss should the container fail. However, there are files like my themes and images that I want to make sure I don't lose if the container decides to restart or when I put in a customer container and CI/CD. Enter -> Azure Storage and Path mappings in Web App for Containers!

I created a File Share in my storage account called contentfiles, where I can upload/update this content as well as interact with it using any SDK or even an Azure Function.

Next, to map the path to the container set it in the Path mappings section of my configuration.

Path Mappings

Be sure to set the ENV in configuration

paths__contentPath to /var/lib/ghost/content_files

This can also be done using the Azure CLI, however I found it easier using the GUI for this task.

az webapp config storage-account add -g <resource-group> --custom-id Content-name [Storage Account Name] --share-name contentfiles --access-key [Storage Account Key] mount-path /var/lib/ghost/content_files

Other configuration settings

NODE_ENV : production
url : .azurewebsites.net OR your custom domain (cause links to break)
WEBSITES_PORT : 2368 (internal port for ghost) - this my be a regression for auto detecting port used by containers.
PORT : 2368 (internal port for ghost) - this my be a regression for auto detecting port used by containers. not a duplicate

Additional Notable changes

I run Cloudflare for DNS and caching, so there were some changes I needed to make for my custom domains. The DNS changes took about 30 seconds to make the A, CNAME and TXT updates. Either it's much easier or I know what I'm doing.

Caching was purged, pushed a button and SSL - done, comes free with Cloudflare.

After all that, very happy. Was great and Azure was super easy to get setup on containers and now I am positioned for CI/CD on GitHub Actions or Azure DevOps or whatever I want. Bring on the updates!

So Many Ways to Learn .NET in 2020

Shayne Boyer — Wed, 08 Jan 2020 16:59:46 +0000

It's a new year and sure to be on your list, as well as mine, is learning. Here is a quick list of some .NET learning resources for you to check off that resolution for 2020.

.NET 101 Videos - dot.net/videos

In September of 2019, many members if the .NET team created over 100 videos for topics including C# 101, .NET, desktop apps, ASP.NET web apps, learning containers and Docker. More have been added recently including a whole series around creating and managing NuGet packages.

All of the videos are available on YouTube or Channel 9.

Microsoft Learn

Microsoft Learn allows you to master core concepts at your speed. Take 15 minutes to complete a module or start a whole learning path over a course of an hour or multiple hours on your schedule.

C# Modules

Over 20 Modules on C# available.

ASP.NET / .NET Core Modules

Create a web API with ASP.NET Core - Create a RESTful service with ASP.NET Core that supports Create, Read, Update, Delete (CRUD) operations.

Create a web UI with ASP.NET Core - ASP.NET Core supports creating webpages using a built-in templating engine called Razor. In this module, you'll learn how to create webpages using Razor with ASP.NET Core.

Secure an ASP.NET Core web app with the Identity framework - Add authentication and authorization to an ASP.NET Core web app using the Identity framework.

Persist and retrieve relational data with Entity Framework Core - Query and manage relational data in a .NET Core app with an Object-Relational Mapper.

Docs, Docs, Docs!

There are so many pages of tutorials, reference pages, SDK, and examples in the docs. https://docs.microsoft.com/ is the entry point, however here are the .NET related pages for you.

Others

On.NET Show - Channel 9 show hosted by Cecil Phillip cecilphillip, Rich Lander @runfaster2000 and other guest hosts. Great content where they look at the APIs, the tooling, and some of the interesting projects being created in the .NET ecosystem.
Azure Friday - Join Scott Hanselman, Donovan Brown, or Lara Rubbelke as they host the engineers who build Azure, demo it, answer questions, and share insights.
Cloud Native Show - Follow Shayne as he explores Microsoft campus to get the definition (and more) straight from engineers and product teams.

This is in no way an exhaustive list, there are many blogs, and other video resources out on the web. Feel free to comment with your favorites!

Where does serverless fit in cloud native apps?

Shayne Boyer — Mon, 09 Dec 2019 16:15:38 +0000

A critical characteristic of cloud native architectures it the ability to dynamically scale and support massive numbers of users, events, and requests on these distributed applications and teams. Serverless can help in handling these challenges without having to worry about infrastructure or even an entire application. Jeff jumps into the show production and chats with us on how this fits into the plan for cloud native apps.

Topics from the show

Azure Functions
Azure Logic Apps
Azure Application Insights

See more episodes at https://aka.ms/cloudnativeshow or on our YouTube Channel!

.NET Conf

Shayne Boyer — Tue, 27 Aug 2019 23:14:34 +0000

.NET Conf is a FREE, 3-day virtual developer event co-organized by the .NET community and Microsoft. This year .NET Core 3.0 will launch at .NET Conf 2019! Come celebrate and learn about the new release.

Download the .NET Core 3.0 Preview

Find out more at www.dotnetconf.net

Listen to Beth and Golnaz discuss the event, some history and more.

More .NET Resources

Have you tried ML.NET?

Shayne Boyer — Sun, 25 Aug 2019 20:37:04 +0000

ML.NET is an open source and cross-platform machine learning framework made for .NET developers.

Using ML.NET you can easily build custom machine learning models for scenarios like sentiment analysis, price prediction, sales forecasting, recommendation, image classification, and more.

ML.NET 1.0 was released at //Build 2019, and since then the team has been working hard on adding more features and capabilities.

Through the survey below, we would love to get feedback on how we can make your journey to infuse Machine Learning in your apps easier with .NET.

Help shape and improve ML.NET for your needs by taking the short survey below!

Take the survey!

Originally posted by Bri Achtman at Hey .NET! Have you tried ML.NET?

DEV Community: Shayne Boyer

🚀 Build Your First Remote MCP Server on Azure Using TypeScript — and Supercharge Your AI Workflows

🌩️ Why Build a Remote MCP Server?

📘 Start With the Official Microsoft Docs

👉 Full Tutorial (TypeScript)

👉 Quickstart: Custom Remote MCP Server

👉 Official Sample Repo

🛠 How It Works (Simplified)

💡 What Developers Are Building Today

🔹 Cloud & infra automation

🔹 DevOps workflows

🔹 Secure enterprise data connectors

🔹 Knowledge tools

🔹 Multi-step orchestration

🌐 Why TypeScript + Azure Functions Is a Perfect Match

🧪 Try It Yourself — And Give Feedback

🙌 Wrap-Up

Building Container Images Without Docker: Introducing pycontainer-build

The Problem with Traditional Container Builds

Enter pycontainer-build

How It Works

Real-World Example

Key Features

🎯 Zero Docker Dependencies

🧠 Smart Auto-Detection

📦 Base Image Support

🚀 Registry Integration

⚡ Layer Caching

🔒 Security & Compliance

🔌 Toolchain Integrations

Use Cases

Configuration Options

Command-Line Interface

Configuration File (pycontainer.toml)

Python API

Under the Hood: OCI Image Spec

Comparison with Other Tools

Current Limitations

Getting Started

Installation

Quick Example

Push to GitHub Container Registry

Why This Matters

For Python Developers

For DevOps Teams

For the Python Ecosystem

Roadmap

Try It Out and Share Feedback!

Acknowledgments

Unveiling VS Code for Web - Azure

Why Choose VS Code for Web - Azure?

One-click Provision and Deploy

The Power of Azure on Your Browser

Saving Your Work

Conclusion

References

Streamline Your Azure Workflow with GitHub Copilot for Azure in VS Code

Why GitHub Copilot for Azure?

Key Features

Get Started Now!

.NET Learning Challenge!

Compete, Learn, and Develop Skills

ASP.NET Community StandUp : FAST Framework

Links from the show

Are you reading the "What's New" in .NET docs?

Updating the my blog from AppService to Containers...finally

Choosing the setup

Analytics

Azure Web App for Containers

MySQL

Other configuration settings

Additional Notable changes

So Many Ways to Learn .NET in 2020

.NET 101 Videos - dot.net/videos

Microsoft Learn

C# Modules

ASP.NET / .NET Core Modules

Docs, Docs, Docs!

Others

Where does serverless fit in cloud native apps?

Configuration File (`pycontainer.toml`)