Jenkins Pipeline Environment Variables

Welcome to the first blog post of the "Jenkins Pipeline Cookbook" series. Today we are focus on working with Jenkins Pipeline environment variables effectively. You will learn how to define env variables, how to updated them, and how to use them in boolean expressions correctly.

You can jump directly to...

List environment variables
- Using env-vars.html
- Using shell command
Reading environment variables
Setting environment variables
Overriding environment variables
Storing Boolean values in environment variables
Capturing sh command output in the env variable

List environment variables

Let’s start by listing all available environment variables. Two different ways allow you to do so.

Using `env-vars.html`

You can open ${YOUR_JENKINS_HOST}/env-vars.html page on your Jenkins master server to get a list of all environment variables listed on an HTML page.

Using shell command

Alternatively, you can list all environment variables by executing the printenv shell command.

Listing 1. Jenkinsfile

pipeline {
    agent any

    stages {
        stage("Env Variables") {
            steps {
                sh "printenv"
            }
        }
    }
}

It might be useful to use printenv | sort command combination to get a sorted list of env variables.

Reading environment variables

You can access environment variables in pipeline steps through the env object, e.g., env.BUILD_NUMBER will return the current build number. You can also use a shorthand version BUILD_NUMBER, but in this variant may be confusing to some users - it misses the context that the BUILD_NUMBER comes from the environment variable.

Listing 2. Jenkinsfile

pipeline {
    agent any

    stages {
        stage("Env Variables") {
            steps {
                echo "The build number is ${env.BUILD_NUMBER}"
                echo "You can also use \${BUILD_NUMBER} -> ${BUILD_NUMBER}"
                sh 'echo "I can access $BUILD_NUMBER in shell command as well."'
            }
        }
    }
}

Setting environment variables

The environment variables can be set declaratively using environment { } block, imperatively using env.VARIABLE_NAME, or using withEnv(["VARIABLE_NAME=value"]) {} block.

Listing 3. Jenkinsfile

pipeline {
    agent any

    environment {
        FOO = "bar"
    }

    stages {
        stage("Env Variables") {
            environment {
                NAME = "Alan"
            }

            steps {
                echo "FOO = ${env.FOO}"
                echo "NAME = ${env.NAME}"

                script {
                    env.TEST_VARIABLE = "some test value"
                }

                echo "TEST_VARIABLE = ${env.TEST_VARIABLE}"

                withEnv(["ANOTHER_ENV_VAR=here is some value"]) {
                    echo "ANOTHER_ENV_VAR = ${env.ANOTHER_ENV_VAR}"
                }
            }
        }
    }
}

Overriding environment variables

Jenkins Pipeline supports overriding environment variables. There are a few rules you need to be aware of.

The withEnv(["env=value]) { } block can override any environment variable.
The variables set using environment {} block cannot be overridden using imperative env.VAR = "value" assignment.
The imperative env.VAR = "value" assignment can override only environment variables created using imperative assignment.

Here is an exemplary Jenkinsfile that shows all three different use cases.

Listing 4. Jenkinsfile

pipeline {
    agent any

    environment {
        FOO = "bar"
        NAME = "Joe"
    }

    stages {
        stage("Env Variables") {
            environment {
                NAME = "Alan" // overrides pipeline level NAME env variable
                BUILD_NUMBER = "2" // overrides the default BUILD_NUMBER
            }

            steps {
                echo "FOO = ${env.FOO}" // prints "FOO = bar"
                echo "NAME = ${env.NAME}" // prints "NAME = Alan"
                echo "BUILD_NUMBER =  ${env.BUILD_NUMBER}" // prints "BUILD_NUMBER = 2"

                script {
                    env.SOMETHING = "1" // creates env.SOMETHING variable
                }
            }
        }

        stage("Override Variables") {
            steps {
                script {
                    env.FOO = "IT DOES NOT WORK!" // it can't override env.FOO declared at the pipeline (or stage) level
                    env.SOMETHING = "2" // it can override env variable created imperatively
                }

                echo "FOO = ${env.FOO}" // prints "FOO = bar"
                echo "SOMETHING = ${env.SOMETHING}" // prints "SOMETHING = 2"

                withEnv(["FOO=foobar"]) { // it can override any env variable
                    echo "FOO = ${env.FOO}" // prints "FOO = foobar"
                }

                withEnv(["BUILD_NUMBER=1"]) {
                    echo "BUILD_NUMBER = ${env.BUILD_NUMBER}" // prints "BUILD_NUMBER = 1"
                }
            }
        }
    }
}

Storing `Boolean` values in environment variables

There is one popular misconception when it comes to using environment variables. Every value that gets stored as an environment variable gets converted to a String. When you store boolean’s false value, it gets converted to "false". If you want to use that value in the boolean expression correctly, you need to call "false".toBoolean() method.^[1]

Listing 5. Jenkinsfile

pipeline {
    agent any

    environment {
        IS_BOOLEAN = false
    }

    stages {
        stage("Env Variables") {
            steps {
                script {
                    if (env.IS_BOOLEAN) {
                        echo "You can see this message, because \"false\" String evaluates to Boolean.TRUE value"
                    }

                    if (env.IS_BOOLEAN.toBoolean() == false) {
                        echo "You can see this message, because \"false\".toBoolean() returns Boolean.FALSE value"
                    }
                }
            }
        }
    }
}

Capturing `sh` command output in the env variable

You can also capture output of a shell command as an environment variable. Keep in mind that you need to use sh(script: 'cmd', returnStdout:true) format to force sh step^[2] to return an output so it can be captured and stored in a variable.

Listing 6. Jenkinsile

pipeline {
    agent any

    environment {
        LS = "${sh(script:'ls -lah', returnStdout: true)}"
    }

    stages {
        stage("Env Variables") {
            steps {
                echo "LS = ${env.LS}"
            }
        }
    }
}

If you want to learn more about Jenkins and CI/CD pipelines, check out "Continuous Delivery with Docker and Jenkins" book by Rafal Leszko. This book will guide you through the process of setting up continuous delivery pipelines using Jenkins and Docker. Highly recommended!

1. Read more about "The Groovy Truth" here - https://groovy-lang.org/semantics.html#_strings

2. https://jenkins.io/doc/pipeline/steps/workflow-durable-task-step/#sh-shell-script

Thank you!

If you read up to this point - leave me a comment below. It means a lot to me to hear your opinion. Share this blog post on Twitter, Facebook, or LinkedIn, if you found it helpful.

I will be very grateful if you consider supporting my work and effort. I run this blog to share my experience, inspire, and help other people to become better problem solvers.

List environment variables

Using env-vars.html

Using shell command

Reading environment variables

Setting environment variables

Overriding environment variables

Storing Boolean values in environment variables

Capturing sh command output in the env variable

Thank you!

You may also like

Building Java and Maven docker images using parallelized Jenkins Pipeline and SDKMAN!

Jenkins Scripted Pipeline vs. Declarative Pipeline - the 4 practical differences

Using SDKMAN! as a docker image for Jenkins Pipeline - a step by step guide

Using `env-vars.html`

Storing `Boolean` values in environment variables

Capturing `sh` command output in the env variable