Understanding Continuous Integration, Delivery, and Deployment with a Step‑by‑Step VSTS CI/CD Pipeline and Docker

What are CI, CD, and Continuous Deployment? CI means that after developers push new code, the system immediately builds and runs unit tests. CD adds automatic deployment of the built artifact to a production‑like environment (e.g., a staging server). Continuous Deployment further automates the final push to the production environment.

More concepts can be found at the linked cheat‑sheet.

Why use CI/CD? It reduces repetitive work and improves efficiency. Instead of installing Jenkins locally, the tutorial uses Microsoft VSTS (Azure DevOps) for a lightweight, cloud‑based solution.

Resources required

A Tencent Cloud Ubuntu server with Docker installed.

A VSTS account to create a free Git repository (up to 5 users).

Step 1 – Create a VSTS project

Step 2 – Create a simple console web application

Install the required NuGet packages.

Note: the application listens on *:5001 (not localhost:5001 ) so Docker can reach it.

Step 3 – Configure VSTS CI

Set up a trigger so that any push to the master branch starts a build.

Select the ASP.NET Core template; ensure the build runs on a Linux agent.

Remove the unnecessary test task.

Step 4 – Configure Release

Add an artifact that downloads the CI output into a folder named _test_netcore-docker .

Create a production environment (empty process) that will run on the Tencent Cloud server via SSH.

Add an SSH task to copy the published files, unzip them, and generate a Dockerfile.

Shell commands used in the task (wrapped in tags):

unzip -q site/drop/TestDotnetcore.zip -d site/drop/

sudo rm -rf site/drop/*.zip

touch site/drop/Dockerfile

echo "FROM microsoft/aspnetcore:2.0" >> site/drop/Dockerfile

echo "COPY . /publish" >> site/drop/Dockerfile

echo "WORKDIR /publish" >> site/drop/Dockerfile

echo "EXPOSE 5001" >> site/drop/Dockerfile

echo "CMD [\"dotnet\", \"TestDotnetcore.dll\"]" >> site/drop/Dockerfile

sudo docker build --rm -t test_image -f site/drop/Dockerfile site/drop/

sudo docker ps -q --filter "name=test_netcore" | grep -q . && sudo docker rm -f test_netcore || true

sudo docker run --name test_netcore -d -p 5001:5001 test_image:latest

if sudo docker images -f "dangling=true" | grep ago --quiet; then sudo docker rmi -f $(sudo docker images -f "dangling=true" -q); fi

sudo rm -rf site/drop

After committing a change to master , the CI pipeline builds the project, the Release pipeline creates the Docker image, runs the container on the cloud server, and the application becomes reachable on port 5001.

The final Docker image test_image and container test_netcore are verified to be running successfully.

Original article: https://www.cnblogs.com/yudongdong/p/9042750.html