How to Run Command-Line Scripts

A build configuration in TeamCity is a job. When you set up a project, TeamCity will scan the files in your repository and autodetect build steps for you.

In this tutorial, we won’t continue with the autodetected build steps because we have our own command-line script to run. However, it’s worth noting that TeamCity can scan a project’s source VCS repository and autodetect suitable build steps in Node.js, Kotlin, Python, Ant, NAnt, Gradle, Maven, MSBuild, Visual Studio solution files, PowerShell, Xcode project files, Rake, and IntelliJ IDEA projects.

To continue with manual build step configuration, click this link:

Since we want to run a command-line script, let’s choose Command Line from the drop-down menu.

In the next step, let’s choose the Run Custom Script option from the drop-down menu and paste our build script in the appropriate field. We’re going to run the mvn clean package command. Just like many other tools, all our build agents have this package installed on them by default.

mvn clean package will compile the Java sources, run some tests, and create a .jar file.

In the deployment step, we’re running the AWS command-line tool, and then copying that .jar file to a private S3 bucket. Then we just have to hit Save.

Before we hit the Run button, we need to specify the credentials for our AWS S3 bucket: an access key ID and the access key secret. One of the ways you can do this is by using Parameters in TeamCity.

To do that, click Parameters | Add New Parameter. For now, we’re going to go with one environment variable. Enter the name, which is AWS_ACCESS_KEY_ID. You’ll also need to select the kind of the new parameter – Environment variable (env.) in our case – and paste the actual access key ID.

Please note: The access key that we’re using is only for demonstration purposes, so you won’t be able to reuse this exact key.

We also want to edit the parameter and set the type to Password. This means that TeamCity will hide the parameter value not just in the UI, but also in log messages and everywhere else.

You will then see that the access key ID is already masked.

Similarly, we can add another environment variable: the secret access key. After that, we should be able to run the build.

To do that, click the Run button which will redirect you to the overview page. While TeamCity is running the build, it will display the build logs, explaining what’s happening in every detail.

Once your build is finished running, TeamCity will display the build status and data, such as when the build was triggered and by whom, how long it took to run, and what build agent was used.

Because we’re using command-line scripts, the only effective output we have is the build log. If we decided to use a specific TeamCity runner instead (e.g. the Maven runner), we’d also get a test report, coverage report, and many other features, such as parallel tests – all out of the box.

To access the build log, click the Build Log tab in your build results output page. TeamCity makes reading the logs easy! Instead of forcing you to download the full build log onto your machine (which you can still do) and then opening it with Notepad++, TeamCity will collapse the build log and make it easily searchable with Ctrl+F.

That’s all for now! Watch other tutorials to learn how to make use of TeamCity-specific features, like test reports or code coverage reports.

Happy building!

For more information, see TeamCity's command line documentation.