Description of problem:

When some systemd services in the agent ISO (e.g. start-cluster-installation.service) are stopped, the code in them continues to run due to KillMode: none being set in the systemd service. This can cause confusing logs like:

Mar 13 09:36:18 master-0 start-cluster-installation.sh[13477]: jq: error (at <stdin>:1): Cannot index string with string "status"

Steps to Reproduce:

1. Stop assisted-service.service after cluster installation has started

Actual results:

Observe that the script from start-cluster-installation.service is still running:

journalctl -f -u start-cluster-installation.service

Expected results:

Scripts in systemd units that get stopped are stopped.