In distributed processing, the distribution of activity across two or more servers in order to avoid overloading any one with too many requests from users.