Re-run scripted pipeline when agent loses connection with master

[Nahue]

I'm trying to configure something that its harder than I thought.

There is a possibility that I can lose an agent in the middle of the pipeline (this is not a problem, is expected due my k8s configuration, its not ofen but can happen) so I would like to re-run the pipeline when this happens.

Due is a pipeline Naginator and Periodic Reincarnation plugin will not work, they work with freestyle projects, not with pipelines.

So, what thought was add a post step that will re-run the pipeline if notices that the problem was agent's disconnetion.

I've noticed that, when I kill the agent, the pipeline aborts so this works:

#!/usr/bin/env groovy

pipeline {

  agent {

    label 'agent1'

  }

  stages {

    stage('Verify template') {

      steps {

        container('container1') {

          script{

            sh 'uname -a'

            sh 'sleep 600'

            sh 'date'

          }

        }

      }

    }

  }

  post {

    aborted { 

      build job: "$env.JOB_NAME", wait: false

    }

  }

}

Looks ok, works as expected but also is re-running the pipeline if I abort the pipeline manually (which is expected too).

Other think I noticed is that always is showing Agent was removed in the pipeline's output log, so if I can analyze the output, I would be able to just re-run when I see that message and here is the problem, I don't find a way to read the log from post.

Do you have some clue how to do it? There is a better way?

Thanks!