We're going to restart aprpid-memcached, one node at the time. The next node will be restarted after the previous node is restarted and in already available state.JIRA Ticket: https://29022131.atlassian.net/browse/TOSD-2515