Tag: Azure

Posted on by — Leave a comment

Dynamically checkout github repository based on parameter

In some cases, the DevOps team may need to checkout multiple github repositories in the current working directory of the pipeline. By default the pipeline will download the code of the repository where it belongs.

The below pipeline will download a github repository based on a parameter that user provides. In order to accomplish that you will need a service connection with your github account/organization.

Then using the below pipeline you can download a repository with a specific branch by providing the repository name.

trigger:
- main

pool:
  vmImage: ubuntu-latest

parameters:
- name: repoparam
  type: string
  default: reponame
- name: branchparam
  type: string
  default: main

variables:
  - name: repositoryvar
    value: ${{parameters.repoparam}}
  - name: branchvar
    value: ${{parameters.branchparam}}


resources:
  repositories:
    - repository: gitrepo
      type: github
      name: geralexgr/$(repositoryvar)
      ref: $(branchvar)
      endpoint: geralexgr

steps:
- checkout: gitrepo
  displayName: download repository

For example given that I have created a repository geralexgr/terraform-az-lin-win I will provide only the name of it terraform-az-lin-win and the main branch.

When running the pipeline the selected repository will be downloaded into C:agent/work/buildID/s or /home/vsts/work/buildID/s

Posted on by — 4 Comments

Start multiple VMs using parameters with Azure DevOps pipeline

In this article I will explain how one can start multiple VMs on Azure using a pipeline automation. The steps you need to follow:

First create a service connection with your subscription. You should navigate to service connections -> azure resource manager and then select service principal (automatic)

You will then have to select the scope* the subscription or resource group this service connection will access.

In my scenario I added a friendly name of the subscription as Azure Subscription Service Connection

trigger: none
pr: none
parameters:
– name: vms
type: object
default: ["ubuntu1","windows-1"]
pool:
vmImage: ubuntu-latest
jobs:
– job: startvmjob
displayName: Start VMs
steps:
– checkout: none
– ${{ each vm in parameters.vms }}:
– task: AzureCLI@2
displayName: starting vm ${{vm}}
inputs:
azureSubscription: 'Azure Subscription Service Connection'
scriptType: 'pscore'
scriptLocation: 'inlineScript'
inlineScript: |
$vmrequest = az vm list | ConvertFrom-Json | Where-Object {$_.Name -Match "${{vm}}"}
az vm start –resource-group $vmrequest.resourcegroup –name $vmrequest.name
view raw start-vms-azure-pipeline hosted with ❤ by GitHub

When you run the pipeline you will need to provide the VM names. You can also run this pipeline on a schedule using a cron task. My vms are ubuntu1, windows-1

By running the pipeline

Then by navigating on Azure you can notice the VM as started.

Video tutorial on YouTube:

Posted on by — Leave a comment

Authenticate to Azure management API using az cli

A quick an easy way to authenticate on Azure management API is to use az cli get-access-token command and use this as a Bearer token on your API calls.

Given that you want to access https://management.azure.com/subscriptions/api you can use:

$token =  az account get-access-token | ConvertFrom-Json

You can then get the access token with $token.accesstoken

Finally you can use this token as a Bearer token on your API calls

$headers = @{ Authorization = "Bearer $token.accesstoken" }
Posted on by — Leave a comment

Azure DevOps agent cannot checkout GitHub repository

Recently I faced an issue with my Azure DevOps self hosted container agents. They could not checkout the git repositories and the build stopped due to the default timeout of 60 minutes per run.

This happened for multiple builds, as a result I had to investigate the reason behind this error.

By checking the logs inside the container on the C:\agent\_diag folder I could recognize an error message like the below:

A session for this agent already exists.
Agent connect error: The task agent xxx already has an active session for owner xxx.. Retrying until reconnected.

By searching online, I figured out that this is a reported bug on previous agents versions. In order to resolve, I updated and reconfigured the agent. You can update the agent, either from the GUI or by creating a new container and installing the latest version of azure devops agent.

In order to reconfigure the agent I first took an interactive shell on it.

docker exec -it agent-name powershell.exe

Then inside C:\agent run the below commands.

Reconfigure the agent

This is a temporary fix for your agent. If the problem persists you should open a support ticket on Microsoft to troubleshoot the issue.