将数据（变量/参数）从一个下游作业传递到上游作业，以便将数据传递到另一个下游作业

Question

I have a scenario where I am triggering two downstream jobs one after another sequentially from an upstream job.

I need to return data xyz = 3.1416 (parameter/variable) generated in the first downstream job (Job A) back to the upstream job or read data xyz (parameter/variable) generated in the first downstream job (Job A) from the upstream job.

I want to do that as the upstream job needs to pass this data to the other downstream job (Job B).
All these jobs are pipeline jobs. I am writing the upstream job as an abstraction layer and to automate the trigger of the 2 downstream jobs sequentially one after another.

structure / flowchart of jobs

Answer 1

There are a couple of approaches trying to solve that problem. Calling jobs up and downstream isn't a good idea, because you can create a circular reference between them (ie: A calls B, that calls A again, that calls B...). Your Jenkins probably won't break because it is limited by the number of workers, but still...

Solution A: Use artifacts

You can store your values in JSON or YAML files and then create Jenkins artifacts using the archiveArtifacts() step and the Copy Artifact plugin. That way, jobs and builds can share information amongst them.

Solution B: Use buildVariables

There's a way to downstream jobs return values back to upstream jobs, using a resource known as buildVariables. Here's the code from the upstream job:

def ret = build job: 'downstream_job'
print "The returned value from the triggered job was ${ret.buildVariables.RETURNED_VALUE}"

And in the downstream job:

environment {
    RETURNED_VALUE = ""
}

stages {
    stage('Doing something') {
        steps {
            script {
                print("Hi, I was triggered!")
                env.RETURNED_VALUE = "Blah blah blah"
            }
        }
    }
}

buildVariables can access any environment variable from the downstream job, except build parameters.

Best regards.