Power Platform <> DevOps ALM

(3) Share

Report

Posted on by AlexJohnH

I'm currently trying to implement ALM within Power Platform - leveraging Power Platform Build Tools & DevOps Pipelines.

I've got it all working - having an input parameter to enter the solutionName then exporting the solution, unpacking it, committing it to source control and deploying it upstream. My question is to the people who have this deployed and work with ALM - how do you handle multiple solutions within Power Platform?

As we have a solution model similar to the below:

010 - Web Resources

020 - Choices

030 - Security Roles

040 - Tables

050 - Processes

060 - Plugins

070 - Plugin Steps

080 - Flows

090 - Dashboards

100 - Model Driven Apps

I wouldn't want to have to run a DevOps pipeline 10 times and enter the solutionName parameter each time?

Categories:

Governance & administering

I have the same question (0)

All responses (5)

Answers (0)

Suggested answer

crisfervil 126 on at

Like (0)

Report
Hi Alex,

The options I can think of:

1. Export all the solutions on every run of the pipeline. Simplest. The solutions will always be imported in the right order.

2. Group similar solutions together, and have a pipeline that exports all the solutions of the group every time. For example, Tables and Choices will most likely go together.

3. Configure your pipeline with a parameter type number. Each solution will have a base 2 number associated, for example:

1 - Web Resources

2 - Choices

4 - Security Roles

8 - Tables

16 - Processes

32 - Plugins

If the user wishes to export Tables and Plugins, then will input 8+32 = 40 in the parameter.

Your pipeline will have to check the number input, and decide what solutions to deploy.

I vibe coded the yaml code for the last otion. This is the result:

# azure-pipeline.yaml trigger: none parameters: - name: Solutions type: number default: 2 values: - 1 - 2 - 4 - 8 - 16 - 32 jobs: - job: EvaluateSolutions displayName: 'Evaluate Selected Solutions' steps: - ${{ if eq(parameters.Solutions & 1, 1) }}: - script: echo "Solution 1" displayName: 'Export Solution 1' - ${{ if eq(parameters.Solutions & 2, 2) }}: - script: echo "Solution 2" displayName: 'Export Solution 2' - ${{ if eq(parameters.Solutions & 4, 4) }}: - script: echo "Solution 4" displayName: 'Export Solution 4' - ${{ if eq(parameters.Solutions & 8, 8) }}: - script: echo "Solution 8" displayName: 'Export Solution 8' - ${{ if eq(parameters.Solutions & 16, 16) }}: - script: echo "Solution 16" displayName: 'Export Solution 16' - ${{ if eq(parameters.Solutions & 32, 32) }}: - script: echo "Solution 32" displayName: 'Export Solution 32'

More info: https://learn.microsoft.com/en-us/azure/devops/pipelines/process/runtime-parameters?view=azure-devops&tabs=script&wt.mc_id=DX-MVP-5005348#loop-through-parameters

I hope it helps!

Regards,

Cris

Was this reply helpful? Yes No
AlexJohnH 14 on at

Like (0)

Report

@crisfervil I was trying to initially have boolean parameters in the DevOps Pipeline for each solution and the user could tick each one they'd wish to export at runtime.

But I couldn't get the Power Platform Build Tools Export Solution task to loop through each solution and export them?

Was this reply helpful? Yes No
Suggested answer

crisfervil 126 on at

Like (0)

Report

@AlexJohnH Take a look at the example in the documentation. It shows how to iterate through the parameters.

But if you use boolean parameters, you don't need to iterate through them. Just check each one, and deploy that solution if the value is ticket.

Then, call the Power Platform Build Tools Export Solution task for every solution. Is the same code copied 10 times, with a different solution name.

Was this reply helpful? Yes No
AlexJohnH 14 on at

Like (0)

Report

@crisfervil I did fear that I'd have to have 10 export solution tasks in my pipeline.

So when I come to deploying the solutions to different environments- will I need 10 import tasks as well per environment? I'm deploying to 4 environments, so that would be 40 import tasks which seems slightly crazy...

Was this reply helpful? Yes No
Suggested answer

crisfervil 126 on at

Like (0)

Report

@AlexJohnH you need to separate the deployment process in two: The Pipeline and the Release.

The Pipeline exports the solutions, once per solution. Here you can implement the optimization we mentioned and export only what it was selected by the developer.

At the end of the Pipeline, you need to create Artifacts. The Artifacts are the ouputs of the pipeline and it must contains the zip files for all the exported solutions.

The second part of the story is the Release. The release will take whatever Artifacts are created by the Pipeline, and it will deploy them to every environment you define.

You still need to configure the steps required to deploy to each environment, but once the first is done, the rest is copy and paste job.

More info:

https://learn.microsoft.com/en-us/azure/devops/pipelines/release/releases?view=azure-devops&wt.mc_id=DX-MVP-5005348

https://learn.microsoft.com/en-us/azure/devops/pipelines/artifacts/build-artifacts?view=azure-devops&tabs=yaml&wt.mc_id=DX-MVP-5005348

This is that the release looks like.

I hope it helps!

Cris

Was this reply helpful? Yes No