If you configure both global and client connection policy resource limits, the first limit reached is always honored. For example, if the server-wide maximum concurrent connections limit is reached, then all subsequent connections are rejected until existing connections are closed regardless of whether a client connection policy limit has been reached.