Jenkins pipeline - start pod with image built in previous stage in a different pod - docker

Question

I have a Jenkins deployed on a kubernetes cluster A, and that can spawn pods in the same cluster A. My workflow is the following
Start a pod with an image designed to build a docker image. This image will be built with a deterministic tag example/appname:${GIT_COMMIT}
After the image was build in the previous pod, start a new pod with the image from the previous build (example/appname:${GIT_COMMIT}), and run multiple test tools
Currently I am successfully making a build with
podTemplate(yaml: """
apiVersion: v1
kind: Pod
spec:
containers:
- name: aws-dockerizer
image: example/aws-dockerizer:0.1.7
command: ['cat']
tty: true
volumeMounts:
- name: dockersock
mountPath: /var/run/docker.sock
volumes:
- name: dockersock
hostPath:
path: /var/run/docker.sock
""") {
node(POD_LABEL) {
stage('Clone') {
git url: 'https://github.com/example/my-app/', branch: '${build_branch_name}', credentialsId: 'github-app'
container('aws-dockerizer') {
stage('Build and deploy') {
withAWS(credentials: 'aws-credentials', region: 'eu-central-1') {
sh '''#!/usr/bin/env bash
git config --global --add safe.directory ${WORKSPACE}
scripts/build_and_push_docker_image_to_aws.sh
'''
}
}
}
}
}
}
I'd like to add the following stage to my pipeline. Note that the new pod "experience-${GIT_COMMIT}" CANNOT be started because he image is not available until the previous step is complete.
podTemplate(
label: "experience-${GIT_COMMIT}"
yaml: """
apiVersion: v1
kind: Pod
spec:
containers:
- name: experience
image: example.dkr.ecr.eu-central-1.amazonaws.com/example/appname:${GIT_COMMIT}
command: ['cat']
tty: true
""")
stage('Run tests') {
node("experience-${GIT_COMMIT}") {
stage('Run tests') {
container('experience') {
stage('Run Rspec') {
sh 'bundle exec rspec'
}
post {}
}
}
}
}
}
Any idea if this is possible ? What's the DSL / concepts I need to use ? How do I "merge" the two pieces of code to achieve what I want ?
I've tried to play around a bit, but when I declare both pod templates at the beginning, it hangs the job until the 2nd pod is ready, which it never will be...