Real-World Docker Images für .NET

Docker Images erlauben die standardisierte Verteilung und den homogenen Betrieb von Anwendungen, unabhängig der gewählten Technologie oder Programmiersprache. Jedoch gilt es einiges beim Erstellen von Docker Images zu beachten.

In diesem Webinar zeigt Cloud-Native-Enthusiast und Docker Captain Thorsten Hans, wie Docker Images für reale .NET-Projekte aussehen sollten, um unterschiedliche Aspekte wie effizientes Debugging und die Inner-Loop Performance zu adressieren. Darüber hinaus werden Sie lernen, wie Sie potenzielle Stolpersteine und Schwächen der “Standard”-.NET-Dockerfiles von Microsoft ausmerzen.

Transcript

1. Real-world Docker images for .NET Thorsten Hans @ThorstenHans Cloud-Native Consultant
2. Cloud-Native Consultant @ Thinktecture #Azure #Kubernetes #Cloud-Native #Docker Thorsten Hans thorsten.hans@thinktecture.com thinktecture.com thorsten-hans.com @ThorstenHans
3. Why are you here? (What’s your experience with Docker)
4. What we will cover today • Introduction • Production grade Docker images for .NET Apps • Remote debugging .NET Apps in Docker containers • Inner-Loop Performance for .NET Apps with Docker • Conclusion Talking Points
5. Why are we talking about that • Different requirements must be addressed when building Docker images • Robust and secure for production usage • Troubleshooting and debugging should be easy • Fast feedback is required for inner-loop experience • Sometimes you may want to address all of those requirements using a single Dockerfile Introduction
6. What we will cover today • Introduction • Production grade Docker images for .NET Apps • Remote debugging .NET Apps in Docker containers • Inner-Loop Performance for .NET Apps with Docker • Conclusion Talking Points P R
7. - We’ll start from the default Dockerfile generated by VS Code - We’ll optimize overall image build-time by - Removing unnecessary stages - Streamline build and publish for .NET - Address common pitfalls for running real-world workloads in containers Production grade Docker images for .NET Apps
8. Production grade Docker images for .NET Apps Demo
9. What we will cover today • Introduction • Production grade Docker images for .NET Apps • Remote debugging .NET Apps in Docker containers • Inner-Loop Performance for .NET Apps with Docker • Conclusion Talking Points
10. - VS Remote Debugger is not installed by default - Remote debugger could be - baked into the Docker Image - Installed on demand using docker exec (VSCode does that too, but it’s terrible slow) - Consider strict tagging policies when baking remote debugger into your images to differenciate between debug- and non-debug images Remote debugging .NET Apps in Docker containers
11. Remote debugging .NET Apps in Docker containers Demo
12. What we will cover today • Introduction • Production grade Docker images for .NET Apps • Remote debugging .NET Apps in Docker containers • Inner-Loop Performance for .NET Apps with Docker • Conclusion Talking Points E
13. - As developers, we want fast feedback to improve our overall productivity - Hot-Reload is provided by dotnet watch - Rude edits will result in the app process being restarted - Set DOTNET_WATCH_RESTART_ON_RUDE_EDIT=true to force restart also when rude edits happen Inner-Loop Performance for .NET Apps with Docker
14. Inner-Loop Performance for .NET Apps with Docker Demo
15. What we will cover today • Introduction • Production grade Docker images for .NET Apps • Remote debugging .NET Apps in Docker containers • Inner-Loop Performance for .NET Apps with Docker • Conclusion Talking Points
16. What we’ve covered • You’ve seen how to • optimize “OOB-Dockerfiles” for .NET Apps • add VS Remote Debugger to Docker images for .NET • optimze your inner-loop performance when building .NET Apps Conclusion
17. What we’ve covered • Identify and address different requirements as early as possible • Optimize your Docker images to stay productive in the inner-loop • Lint your Docker images to meet containerization best practices (e.G. dockle) • Harden your Docker images and check for vulnerabilities (docker scan) Conclusion
18. Q&A If you have further questions: shoot me a mail at thorsten.hans@thinktecture.com / thorsten.hans@outlook.com or tweet at @ThorstenHans