Docker is just as great for existing apps as for new microservices. And it is easy to get started: You can simply package your own apps in Docker without changing the source code. Then you can run them locally to verify the solution works and use CI/CD to deploy to a cluster using the exact same artifacts that you use in dev. This is a great way to make the move to the cloud. In this session I’ll demonstrate that process, taking an n-tier application on my laptop and moving it to Azure.

1. Power your move to the cloud
pieterd@docker.com

2. Moving to the cloud can be challenging
Workloads
Existing apps, microservices, ISV/SI, AI, IOT, serverless
Target platforms
SaaS, PaaS, IaaS, CaaS, FaaS
Clouds
Private, hybrid, public

3. Build, ship and run, any app, anywhere
Containers as a Service
Host Only operating system, joined as cluster
Engine Just a Docker service/daemon
App Packaged in container images
Orchestrator Schedules apps on hosts

4. Containers are great
They separate app and infra (Portability)
They automate app deployment (Agility)
They run on a host with only a Docker engine (Security)
They don’t need a dedicated guest os (Density)

5. Containers are not so great
They won’t fix your organizational challenges
They won’t fix your technical debt
They won’t change monoliths into microservices
They don’t have a GUI
They don’t have state

6. Dockerfile
FROM microsoft/aspnetcore:2.0.5
ARG source
WORKDIR /app
EXPOSE 80
COPY ${source:-bin/Debug/netcoreapp2.0/publish} .
ENTRYPOINT ["dotnet", "myapp.dll"]

7. Multi-stage build

8. Base images
microsoft/windowsservercore
microsoft/iis
microsoft/dotnet
microsoft/aspnet
microsoft/nanoserver:<version/fx/platform>
microsoft/dotnet
microsoft/aspnetcore

9. Docker build & run
docker build . –t <hubid>/<imagename>
imageid
docker run –d –p 80:80 <hubid>/<imagename>
containerid
docker inspect <containerid>
ipaddress

10. Demo: Inner loop

11. Demo: outer loop
Build/CI,
Integrate,
Test
3.
1.
Push
code
Production
environments
Run, Manage
5.
Code
Run
Validate
Debug
Inner-Loop
CD, Deploy
4.
Source Code
Control
(SCC)
2.
6.
Azure
Container
Registry
Azure Monitoring
Outer-Loop

12. Modernize traditional applications
Costly maintenance, unsupported hardware/os
Remember portability, agility, security and density
Package existing apps in containers
Run them on Docker EE with supported OS in Azure
Run more on less, scale up/down, upgrade/rollback, recover
App

13. Typical traditional application (2006)

14. Containerization process
Steps
Windows Server with GUI manually
Windows Server with GUI automated
Windows Server without GUI automated
WindowsServerCore container automated
Use tools like Image2Docker and Docker Application Converter
Optimize

15. Image2Docker
Install-Module Image2Docker
Import-Module Image2Docker
ConvertTo-Dockerfile –local –Artifact IIS
–OutputPath c:i2dpetshop -Verbose

16. MTA Dockerfile optimized
FROM microsoft/aspnet:3.5
COPY . /inetpub/wwwroot
EXPOSE 80
#add-windowsfeature aspnet
#RUN msiexec.exe <your.msi>
#RUN powershell.exe <your.ps1>

17. Docker compose & demo

18. .NET Portability Analyzer
CLI / VS EXT

19. Developer creativity and operational excellence

20. Docker Enterprise Edition platform

21. From poc to prod
Governance
Platform
Pipeline
Applications
Assessment
Go Live
PoC
App in
Prod

22. Next steps
Experiment with Docker4Windows/Mac locally
For production focus on support, security, manageability and choice
Azure (ARM, VSTS, OMS) is great for CaaS
This week: DockerCon US keynotes on Wednesday and Thursday

23. THANK YOU :)
pieterd@docker.com