One solution is to have the members re-register periodically, so that the
coordinator can detect the processes which didn't do so since a while, and
One solution is to have the members re-register periodically, so that the
coordinator can detect the processes which didn't do so since a while, and