Design and operational reliability of a Peer-to-Peer distributed remote access laboratory

The reliability of a system reflects the state of different sub-systems and the interaction between them. Broadly, it is a measure of the probability of the system being usable and available for users at any given point in time. For a remote access laboratory this condition is satisfied when users are able to successfully locate and run an experiment to get the desired results. This paper focuses on the characteristics and components in distributed Peer-to-Peer Remote Access Laboratories (P2P RAL). A model is developed to estimate the reliability and contrast its performance with traditional, centralized RAL systems. It is shown that distributed RAL exhibit a lower reliability than centralized architectures. This is offset however by other affordances of the P2P distributed remote access laboratory enabling multiple replicates of the activity to be hosted separately, thereby improving the reliability. The reliability analysis can extended and used to monitor the performance of the system in a distributed manner and attempt to prevent faults from occurring thus ensuring high reliability of the P2P network.