Restart Jenkins controller and nodes in pipeline run

Question

I have a Jenkins setup looks like this:

• Controller in Debian server, installed and updated via jenkins.io apt repository, but the controller is not accepting jobs.
• 2 nodes, each of them on separate Debian servers, connecting to node as permanent agent setup. I put the agent.jar files into a service, hence I do a systemctl restart jenkins-node it will stop the agent, wget the latest jar file from controller, then start the agent. The jobs are running one of these 2 nodes.

What I would like to achieve to write a pipeline what will do an apt dist-upgrade every day on the controller and in the nodes. This is working fine until a new version of Jenkins is released. When it is, the upgrade process will do a systemctl restart jenkins and then it will crush the running pipeline, no matter the job is not running on the controller. Moreover, if I do a systemctl restart jenkins-node in the agents, then the agent where the build is on going will crush as well. Is there a way to make the pipeline job active during Jenkins service restarts?