An-SPf: as an Alternative Architecture of no Single Point Failure of Scalable Transcoding System Based on Kubernetes

The transcoding process is a complex signal conversion and also a CPU-bound process. A conventional software transcoding process requires a machine with a high CPU core count to achieve a fast transcoding process. That process has a limit because of the vertical scaling limitation on a single machine. In the previous research proposed, Morph System can improve this process by utilizing horizontal scaling using cloud computing architecture, but its system has a flaw which is a single point of failure. So, this is very important to explore how to achieve a faster transcoding process while also providing no single point failure to the system. In this work, we proposed a no single point failure of the transcoding system. The proposed system is built on top of the Kubernetes platform. To compare the performance of the proposed system we used a single node virtual machine (VM) and Morph System. We find that the proposed system provides 1,6 faster than a single VM, but about 0,6 times slower than the Morph system. The system is also able to provide no single point failure under the proposed testing scenario and maintain up to 99,99% of success rate.