Fault detection in distributed transactions based on extended pi-calculus

Distributed transactions gradually become the mainstream mode of data processing. How to avoid the transaction congestion due to network delays and site failures, and how to effectively distinguish between these two faults become the hot issues in the study of distributed transactions. In the fault detection and distinction of the distributed transaction, we develop a hierarchical detection model, which has the advantages of detecting path clearly, having the small number of probe packet, and approximating the actual network topology. The model can effectively solve the problem of excessive traffic load of the coordination to shorten the time of parallel processing for each sub-transaction. On the other hand, this can reduce unnecessary transaction submissions caused by failures through the detection and differentiation of the fault to enhance the reliability and availability of the agreement. We select DπF calculus as a modeling language and the extend DπF calculus through adding the clock operator, which intuitively describes the scenarios of node failure and link failure in distributed transaction failures, and then we can distinguish between these two types of failures and validate them by the bisimulation theory.