Reliability of tree-like multicast overlays caused by nodes’ sudden failures is considered as one of the significant issues for the Internet application-layer media streaming service. In this project, we address this issue by designing a distributed and light-weighted protocol named the instantaneous reliability oriented protocol (IRP). Not at all like the greater part of existing empirical solutions, we initially define the overlay reliability problem formally, and propose a protocol containing a node joining algorithm (IRP-Join), a node preemption algorithm (IRP-Preempt), and a node switching algorithm (IRP-Switch) for reactively building and repairing the overlay, and also proactively keeping up the overlay. With the formal issue introduction, we set up a worldview for taking care of the overlay reliability issue by theoretically proving the effectiveness of our algorithms. Moreover, by comparing IRP with existing solutions via simulation-based experiments and real-world deployment, we show that IRP achieves a better reliability, while incurs fewer structural adjustments on the multicast overlay, thus, providing a superior overall performance.