RPC continue_node_{deploy,clean} are called from a conductor handling
the node, so they don't have to go through the hash ring. This avoids
situation when take over happens in the middle of a deploy/clean step
processing, breaking it.
Eventually, we should stop using RPC for that at all, but that will be
a much more invasive change.
Story: #2008200
Task: #40984
Change-Id: I76293f8ec30d5957b99bdbce5b70e87e8378d135