Using MSBuild tool to get code coverage and configure Azure DevOps pipelines to include code coverage results is an easy task for .NET framework based applications. Azure DevOps (formerly VSTS) contains inbuilt functionality to analyze code coverage files generated and publish results back to VSTS itself. However, it is quite a challenge to get it right and working for .NET Core 2.0 based applications. In this blog post, we’ll cover steps on how to get code coverage results for .NET Core based application using SonarQube and Azure DevOps. Continue reading “Configure Code Coverage for Dotnet Core 2.0 based applications using SonarQube and Azure DevOps”
Using SonarQube extesions from Marketplace for Azure DevOps provides much of the integration functionality between Azure DevOps and SonarQube. Once the build pipeline completes, you can login in SonarQube server and view the code analysis results. Based on the code analysis results against the Quality threshold set or default Quality Gate threshold, it will be assigned a rating. However, there is no way to stop check-in of code, if it fails to passes the Quality Gate criteria. However, we can use some PowerShell and SonarQube Web APIs to do this part for us. In this blog post, we’ll learn steps to do the same. Continue reading “Fail Azure DevOps pipeline if build fail to pass the SonarQube Quality Gate”
After upgrading our SonarQube Community version to the newly released version, our build pipelines hosted in the Azure DevOps (formerly VSTS) started failing with below errors:
“02:14:02.317 ERROR: Caused by: Validation of project reactor failed:
o To use the property “sonar.branch.name”, the branch plugin is required but not installed. See the documentation of branch support: https://redirect.sonarsource.com/doc/branches.html”
We were using the same release branches that we were using earlier and there was no changes made to the build configuration. Continue reading “Azure DevOps build pipelines fails with error: To use the property “sonar.branch.name”, the branch plugin is required but not installed.”
Sometimes it happens that for the organization and maintenance purpose, you would like to keep all code related to one component of a product in one place only. Now that component may be complex component having multiple sub parts like API, UI, Database etc. Again, for the purpose of easy maintenance, you would like it all under one repository, which is not an bad idea. Now, imagine if you have different builds configured for different sub-components which are using this source code as their base and are configured to trigger on some event, say commit. Then for each small and big commit, it will trigger build in each of the configured builds. Ideally, if the change was related to only one sub-component say API, then you would need to trigger only API specific build and not the ones for the UI and/or Database. Continue reading “Using Path filters in Build definition in Azure DevOps / VSTS”
Visual Studio Team Services (VSTS), Microsoft’s application lifecycle management system, is to undergo a major shake-up and rebranding. Instead of a single Visual Studio-branded service, it’s being split into five separate Azure-branded services, under the banner Azure DevOps.
The five components:
- Azure Pipelines, a continuous integration, testing, and deployment system that can connect to any Git repository
- Azure Boards, a work tracking system with Kanban boards, dashboards, reporting
Continue reading “Updated: Microsoft VSTS converted to Azure DevOps”
After making some changes to the build definition of one of the dotnet core repositories, we started observing this error with the SonarQube tasks inside the build definition in VSTS:
When we checked the complete logs, we observed below issue:
In git workflow, the ideal strategy to work is to fork a new branch, make changes and finally merge in the main development branches. Over the time, this results in creation of large number of branches which are not required and becomes stale. Although a branch is just a pointer to an commit and does not require more than 40 bytes of disk space, it can be painful to search a long list of branches and deciding what you want to work on. Also since we humans are not good with creating unique names for branches, they can also result in confusion.
Below are some steps to clean branches from git repository to remove the clutter. Continue reading “Delete Branches in Git using commands”