No not need to write in the ceph.conf and not need to do a reboot.
On a single node will propagate. However, you can run the following command on all the nodes to make sure that the scrub load did on all the nodes.
Bash:
ceph config get osd osd_scrub_load_threshold