Study on High Availability and Fault Tolerance

Availability is one of the most important requirements for modern computing systems. In cloud computing, it is common to use it as a key factor in adopting a cloud service. This paper studies the breakdown in calculating the availability and proposes a conceptual model as middleware in order to speed up the time for detection during a system crash. It defines the maximum acceptable outage time for any given period. Through simulations tests, we verified that the proposed model is able to detect the system crash in sub-seconds and improve the overall availability of the system compared to currently used industry solutions.