Set up Tracing on a Jenkins Pipeline

Compatibility

• Supported Jenkins versions:

  • Jenkins >= 2.346.1

• Manual steps: View manually triggered pipelines

• Queue time: View amount of time pipeline jobs wait in the queue before processing

• Logs correlation: Correlate pipeline spans to logs and enable job log collection

• Infrastructure metric correlation: Correlate pipelines to infrastructure host metrics for Jenkins workers

• Custom spans: Configure custom spans

• Custom pre-defined tags: Configure custom tags and metrics at runtime

• Parameters: Set custom parameters such as default branch name and Git information

• Pipeline failure reasons: Identify pipeline failure reasons

Install the Datadog Agent (optional)

The Datadog Jenkins plugin can either report metrics through the Datadog Agent or directly to Datadog if an API key is provided. If you don’t have a Datadog Agent running on the Jenkins controller instance, Datadog recommends installing it first by following the Agent installation instructions.

If the Jenkins controller and the Datadog Agent have been deployed to a Kubernetes cluster, Datadog recommends using the Admission Controller, which automatically sets the DD_AGENT_HOST environment variable in the Jenkins controller pod to communicate with the local Datadog Agent.

Configure the Agent for logs collection

To enable collecting logs from your jobs, configure the Datadog Agent to listen for logs on a TCP port:

With this configuration, the Agent listens for logs on port 10518.

Install the Datadog Jenkins plugin

Install and enable the Datadog Jenkins plugin v3.3.0 or newer:

1. In your Jenkins instance web interface, go to Manage Jenkins > Manage Plugins.
2. In the Update Center on the Available tab, search for Datadog Plugin.
3. Select the checkbox next to the plugin, and install using one of the two install buttons at the bottom of the screen.
4. To verify that the plugin is installed, search for Datadog Plugin on the Installed tab.

Enable CI Visibility on the plugin

There are several ways you can configure the Datadog Jenkins plugin.

Configure with the Jenkins configuration UI

1. In your Jenkins instance web interface, go to Manage Jenkins > Configure System.
2. Go to the Datadog Plugin section, scrolling down the configuration screen.
3. Select the mode Use the Datadog Agent to report to Datadog.
4. Configure the Agent host.
5. Configure the Traces Collection Port if not using the default port 8126.
6. Mark the Enable CI Visibility checkbox.
7. (Optional) Configure your CI Instance name.
8. Check the connectivity with the Datadog Agent.
9. Save your configuration.

Use this option to make the Jenkins plugin report directly to Datadog without using the Datadog Agent. It requires an API Key.

1. In your Jenkins instance web interface, go to Manage Jenkins > Configure System.
2. Go to the Datadog Plugin section, scrolling down the configuration screen.
3. Select the mode Use Datadog API URL and Key to report to Datadog.
4. For Datadog sites other that datadoghq.com, change the Datadog API URL, Log Intake URL and Webhook Intake URL to point to your Datadog site.
5. Enter a valid Datadog API Key.
6. Mark the Enable CI Visibility checkbox.
7. (Optional) Configure your CI Instance name.
8. Save your configuration.

Configuration-as-code

Configure with Groovy

Use environment variables

Validate the configuration

To verify that CI Visibility is enabled, go to Jenkins Log and search for:

Re/Initialize Datadog-Plugin Agent Http Client
TRACE -> http://<HOST>:<TRACE_PORT>/v0.3/traces

Correlate infrastructure metrics

If you are using Jenkins workers, you can correlate pipelines with the infrastructure that is running them. For this feature to work:

1. Install the Datadog Agent in every Jenkins worker.
2. Set and export a new environment variable called DD_CI_HOSTNAME in every Jenkins worker with the worker hostname.

• It must be the same hostname that the Datadog Agent is reporting in the infrastructure metrics for that worker.
• You can use fixed values or other environment variables as valid values.

export DD_CI_HOSTNAME=my-hostname

If you are using Kubernetes to manage your Jenkins instances, add the DD_CI_HOSTNAME environment variable to the pod that executes the Jenkins job. The value of this environment variable depends on what you are using in your Datadog Agent daemonset when reporting the infrastructure metrics.

This is only required for Jenkins workers. For the Jenkins controller, the infrastructure metric correlation does not require additional actions.

Note: Infrastructure metric correlation is supported since Jenkins Plugin v5.0.0+

Enable job log collection

This is an optional step that enables the collection of job logs.

Enable with the Jenkins configuration UI

Enable with configuration-as-code

Enable with Groovy

Use environment variables

Enable Test Visibility

This is an optional step that enables the collection of tests data using Test Visibility.

See the Test Visibility documentation for your language to make sure that the testing framework that you use is supported.

Enable with the Jenkins configuration UI

1. In your Jenkins instance web interface, go to the job or pipeline that you want to instrument and choose the Configure option.
2. In the General configuration section, tick the Enable Datadog Test Visibility checkbox.
3. Enter the name of the service or library being tested into the Service Name input. You can choose any value that makes sense to you.
4. Choose the languages for which you want to enable tests instrumentation. Some of the languages do not support configuration through the UI. To configure Test Visibility for these languages, follow the manual configuration instructions.
5. Optionally, provide additional configuration settings.
6. Click Save.

Enable with manual tracer configuration

Follow the manual Test Visibility configuration instructions that are specific to your language.

Set the default branch name

To report pipeline results, attach the default branch name (for example, main) to pipeline spans in an attribute called git.default_branch. This is usually done automatically, but in some cases the plugin cannot extract this information because it might not be provided by Jenkins.

If this happens, set the default branch manually using the DD_GIT_DEFAULT_BRANCH environment variable in your build. For example:

pipeline {
    agent any
    environment {
        DD_GIT_DEFAULT_BRANCH = 'main'
        ...
    }
    stages {
        ...
    }
}

Propagate Git information in pipelines without a Jenkinsfile from SCM

The Jenkins plugin uses environment variables to determine the Git information. However, these environment variables may not be available if you are not using a Jenkinsfile in your repository, and you’re configuring the pipeline directly in Jenkins using the checkout step.

In this case, you can propagate the Git information to the environment variables in your build. Use the .each {k,v -> env.setProperty(k, v)} function after executing the checkout or git steps. For example:

pipeline {
  stages {
    stage('Checkout') {
        script {
          checkout(...).each {k,v -> env.setProperty(k,v)}
        }
    }
    ...
  }
}

pipeline {
  stages {
    stage('Checkout') {
      script {
        git(...).each {k,v -> env.setProperty(k,v)}
      }
    }
    ...
  }
}

node {
  stage('Checkout') {
    checkout(...).each {k,v -> env.setProperty(k,v)}
  }
  ...
}

node {
  stage('Checkout') {
    git(...).each {k,v -> env.setProperty(k,v)}
  }
  ...
}

Set Git information manually

The Jenkins plugin uses environment variables to determine the Git information. However, these environment variables are not always set automatically due to dependencies on the Git plugin that is being used in the pipeline.

If the Git information is not detected automatically, you can set the following environment variables manually.

Note: These variables are optional, but if they are set, they take precedence over the Git information set by other Jenkins plugins.

DD_GIT_REPOSITORY (Optional)

The repository URL of your service.
Example: https://github.com/my-org/my-repo.git

DD_GIT_BRANCH (Optional)

The branch name.
Example: main

DD_GIT_TAG (Optional)

The tag of the commit (if any).
Example: 0.1.0

DD_GIT_COMMIT_SHA (Optional)

The commit expressed in the hex 40 chars length form.
Example: faaca5c59512cdfba9402c6e67d81b4f5701d43c

DD_GIT_COMMIT_MESSAGE (Optional)

The commit message.
Example: Initial commit message

DD_GIT_COMMIT_AUTHOR_NAME (Optional)

The name of the author of the commit.
Example: John Smith

DD_GIT_COMMIT_AUTHOR_EMAIL (Optional)

The email of the author of the commit.
Example: john@example.com

DD_GIT_COMMIT_AUTHOR_DATE (Optional)

The date when the author submitted the commit expressed in ISO 8601 format.
Example: 2021-08-16T15:41:45.000Z

DD_GIT_COMMIT_COMMITTER_NAME (Optional)

The name of the committer of the commit.
Example: Jane Smith

DD_GIT_COMMIT_COMMITTER_EMAIL (Optional)

The email of the committer of the commit.
Example: jane@example.com

DD_GIT_COMMIT_COMMITTER_DATE (Optional)

The date when the committer submitted the commit expressed in ISO 8601 format.
Example: 2021-08-16T15:41:45.000Z

If you set only repository, branch and commit, the plugin will try to extract the rest of the Git information from the .git folder.

An example of usage:

pipeline {
  agent any
  stages {
    stage('Checkout') {
      steps {
        script {
          def gitVars = git url:'https://github.com/my-org/my-repo.git', branch:'some/feature-branch'

          // Setting Git information manually via environment variables.
          env.DD_GIT_REPOSITORY_URL=gitVars.GIT_URL
          env.DD_GIT_BRANCH=gitVars.GIT_BRANCH
          env.DD_GIT_COMMIT_SHA=gitVars.GIT_COMMIT
        }
      }
    }
    stage('Test') {
      steps {
        // Execute the rest of the pipeline.
      }
    }
  }
}

Customization

Set custom tags for your pipelines

The Datadog plugin adds a datadog step that allows adding custom tags to your pipeline-based jobs.

In declarative pipelines, add the step to a top-level option block:

def DD_TYPE = "release"
pipeline {
    agent any
    options {
        datadog(tags: ["team:backend", "type:${DD_TYPE}", "${DD_TYPE}:canary"])
    }
    stages {
        stage('Example') {
            steps {
                echo "Hello world."
            }
        }
    }
}

In scripted pipelines, wrap the relevant section with the datadog step:

datadog(tags: ["team:backend", "release:canary"]){
    node {
        stage('Example') {
            echo "Hello world."
        }
    }
}

Integrate with Datadog Teams

To display and filter the teams associated with your pipelines, add team:<your-team> as a custom tag. The custom tag name must match your Datadog Teams team handle exactly.

Set global custom tags

You can configure the Jenkins Plugin to send custom tags in all pipeline traces:

1. In the web interface of your Jenkins instance, go to Manage Jenkins > Configure System.
2. Go to the Datadog Plugin section, scrolling down the configuration screen.
3. Click on the Advanced button.
4. Configure the Global Tags.
5. Configure the Global Job Tags.
6. Save your configuration.

Global tags

A comma-separated list of tags to apply to all metrics, traces, events, and service checks. Tags can include environment variables that are defined in the Jenkins controller instance.
Environment variable: DATADOG_JENKINS_PLUGIN_GLOBAL_TAGS
Example: key1:value1,key2:${SOME_ENVVAR},${OTHER_ENVVAR}:value3

Global job tags

A comma-separated list of regexes to match a job and a list of tags to apply to that job. Tags can include environment variables that are defined in the Jenkins controller instance. Tags can reference match groups in the regex using the $ symbol.
Environment variable: DATADOG_JENKINS_PLUGIN_GLOBAL_JOB_TAGS
Example: (.*?)_job_(.*?)_release, owner:$1, release_env:$2, optional:Tag3

Include or exclude pipelines

You can configure the Jenkins Plugin to include or exclude some pipelines:

1. In the web interface of your Jenkins instance, go to Manage Jenkins > Configure System.
2. Go to the Datadog Plugin section by scrolling down the configuration screen.
3. Click on the Advanced button.
4. Configure the Excluded Jobs.
5. Configure the Included Jobs.
6. Save your configuration.

Excluded jobs

A comma-separated list of Jenkins jobs that should not be monitored. The exclusion applies to all metrics, traces, events, and service checks. Excluded jobs can use regular expressions to reference multiple jobs.
Environment variable: DATADOG_JENKINS_PLUGIN_EXCLUDED
Example: susans-job,johns-.*,prod_folder/prod_release

Included jobs

A comma-separated list of Jenkins job names that should be monitored. If the included jobs list is empty, all jobs that are not excluded explicitly are monitored. The inclusion applies to all metrics, traces, events, and service checks. Included jobs can use regular expressions to reference multiple jobs.
Environment variable: DATADOG_JENKINS_PLUGIN_INCLUDED
Example: susans-job,johns-.*,prod_folder/prod_release

Visualize pipeline data in Datadog

Once the integration is successfully configured, both the Pipelines and Pipeline Executions pages populate with data after pipelines finish.

Note: The Pipelines page shows data for only the default branch of each repository.

Troubleshooting

Enable DEBUG log level for the Datadog Plugin

If you have any issues with the Datadog Plugin, you can set the logs for the plugin in DEBUG level. Using this level you are able to see stacktrace details if an exception is thrown.

1. In your Jenkins instance web interface, go to Manage Jenkins > System log.
2. Click on Add new log recorder button.
3. Type the log recorder name. E.g: Datadog Plugin Logs.
4. Add the following loggers to the list:
   • Logger: org.datadog.jenkins.plugins.datadog.clients -> Log Level ALL
   • Logger: org.datadog.jenkins.plugins.datadog.traces -> Log Level ALL
   • Logger: org.datadog.jenkins.plugins.datadog.logs -> Log Level ALL
   • Logger: org.datadog.jenkins.plugins.datadog.model -> Log Level ALL
   • Logger: org.datadog.jenkins.plugins.datadog.listeners -> Log Level ALL
5. Save the configuration.

You may also want to split the loggers into different log recorders.

Once the log recorders are successfully configured, you can check the logs in the DEBUG mode by accessing the desired log recorder through Manage Jenkins > System log.

If you trigger a Jenkins pipeline, you can search for the message Send pipeline traces in the Datadog Plugin Logs. This message indicates that the plugin is sending CI Visibility data to the Datadog Agent.

Send pipeline traces.
...
Send pipeline traces.
...

The Datadog Plugin cannot write payloads to the server

If the following error message appears in the Jenkins Log, make sure that the plugin configuration is correct.

Error writing to server

1. If you are using localhost as the hostname, try to change it to the server hostname instead.
2. If your Jenkins instance is behind an HTTP proxy, go to Manage Jenkins > Manage Plugins > Advanced tab and make sure the proxy configuration is correct.

HTTP 504

If the HTTP 504 error message appears, make sure that the Jenkins proxy configuration is correct.

Failed to send HTTP request: PUT http://localhost:8126/v0.3/traces - Status: HTTP 504

1. If your Jenkins instance is behind an HTTP proxy, go to Manage Jenkins > Manage Plugins > Advanced tab and make sure the proxy configuration is correct.
2. Check that localhost has been configured in the No Proxy Hosts section.

The Datadog Plugin section does not appear in the Jenkins configuration

If the Datadog Plugin section does not appear in Jenkins configuration section, make sure that the plugin is enabled. To do so:

1. In your Jenkins instance web interface, go to Manage Jenkins > Manage Plugins.
2. Search for Datadog Plugin in the Installed tab.
3. Check that the Enabled checkbox is marked.
4. If you enable the plugin here, restart your Jenkins instance using the /safeRestart URL path.

The CI Visibility option does not appear in the Datadog Plugin section.

If the CI Visibility option does not appear in the Datadog Plugin section, make sure that the correct version is installed and restart the Jenkins instance. To do so:

1. In your Jenkins instance web interface, go to Manage Jenkins > Manage Plugins.
2. Search for Datadog Plugin in the Installed tab.
3. Check that the installed version is correct.
4. Restart your Jenkins instance using the /safeRestart URL path.

The Plugin’s Tracer fails to initialize due to APM Java Tracer is being used to instrument Jenkins.

If this error message appears in the Jenkins Log, make sure that you are using the Jenkins plugin v3.1.0+

Failed to reinitialize Datadog-Plugin Tracer, Cannot enable traces collection via plugin if the Datadog Java Tracer is being used as javaagent in the Jenkins startup command. This error will not affect your pipelines executions.

Further reading

Additional helpful documentation, links, and articles:

Explore Pipeline Execution Results and PerformanceDOCUMENTATION Extend Pipeline Visibility by tracing individual commandsDOCUMENTATION Troubleshooting CIDOCUMENTATION