I'm evaluating Azure Devops and want to have following workflow: I want to have a pipeline build and run it on our self-hosted build agent, but if non them is available I want to fall back on hosted agent and vice versa. According to documentation https://learn.microsoft.com/en-us/azure/devops/pipelines/agents/agents?view=azure-devops in the yaml file you need to define in which pool will job run.
Can I have a pipeline build without having to define a pool/define many pools as fallback?
Related
We have been converting our Release pipelines in Azure DevOps to instead be YAML files that run as Pipelines. This is so we can store our deployment process as code. The deployment process is working well - so a developer commits code and the Pipeline builds and publishes the artifact in one stage and then it auto-deploys to the QA environment in another stage. Subsequent stages deploy to an Environment (e.g. QA, Staging, Production) each of which requires some approval(s). The deployments themselves aren't the issue.
What I'm struggling with is unlike the old Releases there isn't a dashboard that will tell me which version of the project is in each environment. The Pipeline summary represents each stage for the Run as a dot (running, succeeded, failed, canceled, etc.) but wasn't built to represent what each environment has (probably because a stage doesn't have to be a deployment).
Is there somewhere else I can look for this information or do I have to build my own dashboard by calling the AzDO APIs? Looking at the Environment gives a list and I can root through the history, but that's not the experience our developers are looking for.
I have a nodejs web application that I build in Azure Pipelines. I am planning to deploy the generated artifacts on a Azure VM (probably a dev test labs), as part of one of the pipeline steps.
I want to now run browser tests by pointing the browser to the hosted URL in the Azure VM. I want to use the Azure windows and linux VMs in a build pipeline to run the tests on this remote Azure VM and publish the results to the pipeline. These would be karma tests essentially running on the nodejs server.
In my current design, the test results are going to be available on the Azure VM hosting the nodejs application.
What I don't understand is how can I get these test results back to
the Azure Pipeline for publishing the same?
Is there a way I can architect this solution without having to setup my Azure VM as a
pipeline agent in Azure DevOps?
Is there a standard pattern to design such continuous test infrastructure using Azure DevOps?
Thanks
According to your description, you just want to use Microsoft host agent to access an url on your self-host agent (ignore it's Azure VM or your own physical machine, same to host agent).
It depends if that url are accessible through public internet.
The simplest solution here is deploy your build agent on that Azure VM directly. Then run build and test. You can do this through the following script and tasks:
run ng test or any command to raise your tests
publish test results with PublishTestResults task
publish code coverage results with PublishCodeCoverageResults task
Microsoft-hosted agent pool will not work for you with every scenarios. For many teams this is the simplest way to run your jobs. You can try it first and see if it works for your build or deployment. If not, you can use a self-hosted agent. Self-hosted agents give you more control for your builds, tests and deployments.
In your scenario, setup your Azure VM as a pipeline agent and run build/test on it should be the simplest and convenient solution.
I have two pipelines on azure devops that do the following:
BUILD PIPELINE
Asks the user for the desire branch name and ticket id (from our ticketing system). Using two "settable at queue time" variables.
Gets the files from the specified branch name, builds the project and generate the artifacts that are later use by the RELEASE pipeline
RELEASE PIPELINE
This release pipeline is being triggered by the completion of the BUILD pipeline specified before. This pipeline completes the following tasks:
Creates new app service slot using an Azure CLI command task. In this task I need the ticket id, in order to create the slot with the corresponding ticket id as its name
Deploy the application to the created app service slot
I haven't found a way to pass to the RELEASE pipeline the ticket id the user entered when running the BUILD pipeline that triggered this RELEASE.
Unfortunately, there's no such a build-in function in Azure DevOps. While thanks to Martin Hinshelwood, we had an extension- Variable Tools for Azure DevOps Services to achieve the goal.
As you can see in the extension's doc, add a Variable Save Task in your build pipeline and a Variable Load Task in your release pipeline. After these two tasks, the variables are fully passed to your release.
Check below as reference:
I'm building an Azure DevOps pipeline to deploy a custom-build powershell application to several on-prem environments that we support. I configured the required agent pools and installed them as a service on the on-prem environment.
Next, I have set up my pipeline in Azure DevOps, selecting a GitRepo:
Build (with the steps: Use Nuget, Nuget Restore, Build solution, Update version, Copy Files, and Publish build Artifact)
Release (with step: Publish Build Artifact)
Some things are unclear for me:
Do I need the Publish Build Artifacts twice? Can the Build pipeline
end with Copy Files step, and that the Release pipeline picks up this
artifact?
It is my understanding that the release publishes the app to the
on-prem environment (in my case). Where can I set a custom path (ie:
C:\deployed_apps) where the app needs to be deployed? When I tested
this pipeline, I got errors that the path I created using an variable
was not found.
What am I missing in my setup to get this pipeline working?
As #Shayki Abramczyk
pointed out this task is not for deployment, it just upload your Build Artifacts to azure devops server where your release pipeline can download directly.
In your case, if you want deploy your application to several on-prem environments, You need to create a deployment group first, A deployment group is a logical set of deployment target machines that have agents installed on each one. You application will be deployed to those machines in a deployment group in release pipeline. Check here for more details about deployment group.
After the deployment group is created, you can add a deployment group job by click the 3 dots, and then specify your deployment group as below pic shows, You can then simply add a copy file task or other deployment tasks to deploy your application to your on-prem machine.
In the release pipeline you shouldn't use the Publish Build Artifact. in the end of the build you put this step, what this step does? upload your artifacts to the Azure DevOps or to a file share. now in the release pipeline you choose the build artifact (in the left pane). the first thing that the agent does when the release pipeline started is to download the build artifacts to the agent. now you need to take them and deploy it to your environments. how? it depends which kind of application is (it can be just copy files, it can be deploy to IIS, etc.).
You can put the path in the variables tab and use this variable when you deploy the app (with copy files task, for example).
I have created a custom build agent in Azure DevOps. I can view the agent in agent pool, but there is no option to select the agent when queuing the build.
I want to run the build using that custom build agent. So please guide me on how to configure the build pipeline to run with custom build agent.
When you create a self-hosted (I assume you mean this by custom, right?) agent it goes to the Default pool.
So you need to select this pool when you run a build. You can either set this pool as a default for the build in the build configuration OR you can select it right before you start the build instead of the default one.