Disclosed is a system for ensuring the continuity of a proxy-based service, and a method thereof. The system includes a target server that processes a queue of access requests received from a plurality of client devices and processes the queue until a first time point when the waiting amount of access request exceeds a predetermined first threshold, a queue management server that processes the queue from the first time point until a second time point when the waiting amount exceeds a predetermined second threshold, and a mirror server that processes the queue from the second time point.