Create Jenkinsfile from "execute shell" commands - bash

Question

I have a series of Execute shell boxes on a Jenkins build. After 3 days of Google and watching videos I need help. I am more a sysadmin than a coder so I'm having difficulty in creating a Jenkinsfile with the correct options and syntax. Can anyone advise? I need to create a pipeline. Anything in <name> is like that for security reasons, I have the real values in the files.
Execute shell
mkdir -p deploy
Execute shell
cp -R code/api deploy/
cp docker/Dockerfile.dev deploy/
(cd deploy/api/<Name>.<Name>.Web/ && aws s3 cp --recursive --region=eu-west-1 s3://config.<name>/audience-view/atg/dev/API/ .)
Execute shell
cd deploy && docker build -t <name> -f Dockerfile.dev .
Execute shell
aws ecr get-login --region eu-west-1 > docker_login.sh && chmod +x
docker_login.sh && ./docker_login.sh
docker tag <name>:latest 543573289192.dkr.ecr.eu-west-
1.amazonaws.com/<name>:latest
docker push <name>.dkr.ecr.eu-west-1.amazonaws.com/<name>:latest
Execute shell
docker rmi audience-view-dev-api
docker rmi 543573289192.dkr.ecr.eu-west-1.amazonaws.com/<name>:latest
Execute shell
RUNNING_TASKS=$(aws ecs list-tasks --region=eu-west-1 --cluster <name> --family <name> --query 'taskArns')
if [ "$RUNNING_TASKS" != "[]" ]; then
TASK_ARN=$(aws ecs list-tasks --region=eu-west-1 --cluster a<name> --family <name> --query 'taskArns[0]' | sed 's/\"//g')
aws ecs stop-task --region=eu-west-1 --cluster=<name> --task=$TASK_ARN --reason="Deployment from Jenkins"
while [ $RUNNING_TASKS != "[]" ]; do
sleep 5
RUNNING_TASKS=$(aws ecs list-tasks --region=eu-west-1 --cluster <name> --family <name> --query 'taskArns')
done
fi
Execute shell
TASK_ARN=$(aws ecs start-task --region=eu-west-1 --cluster <name> --task-definition <name> --container-instances 5f0c5b75-64a2-45cf-8ced-d6a6d13d2666 --query 'tasks[0].taskArn' | sed 's/arn:aws:ecs:eu-west-1:543573289192:task\///' | sed 's/\"//g')
TASK_STATUS=$(aws ecs describe-tasks --region=eu-west-1 --cluster <name> --tasks $TASK_ARN --query 'tasks[0].lastStatus')
while [ $TASK_STATUS == "PENDING" ]; do
echo $TASK_STATUS
TASK_STATUS=$(aws ecs describe-tasks --region=eu-west-1 --cluster <name> --tasks $TASK_ARN --query 'tasks[0].lastStatus' | sed 's/\"//g')
if [ $TASK_STATUS == "STOPPED" ]; then
echo $(aws ecs describe-tasks --region=eu-west-1 --cluster <name> --tasks $TASK_ARN --query 'tasks[0].containers[0].exitCode')
exit 1
fi
done

Answer 1

Jenkins is best used as the glue to connect all the build pieces together, not the build script itself. As Alfe mentioned, it would be best to have this all in a shell script and then run the shell script with Jenkins.
BUT, if you really want to do this in a Pipeline job, it would look something like this (declarative pipeline):
pipeline {
agent any
stages {
stage('setup') {
steps {
sh "mkdir -p deploy"
}
stage('nextStage') {
steps {
sh """
cp -R code/api deploy/
cp docker/Dockerfile.dev deploy/
(cd deploy/api/<Name>.<Name>.Web/ && aws s3 cp --recursive --region=eu-west-1 s3://config.<name>/audience-view/atg/dev/API/ .)
"""
}
}
stage('anotherStage') {
steps {
echo "repeat for all your shell steps"
}
}
}
}