Virtual Priority Assignment Policy to Distributed Real-Time Transactions

In a distributed environment, tasks generally have processing demands at multiple sites. A distributed task is usually divided into several cohorts, each to be executed in order at some site. This paper deals with an important aspect of distributed real time transaction processing in distributed real-time databases, namely the problem of assigning priorities to transactions. In distributed real time database systems deadline of transactions as well as cohorts affect other transactions during their execution. We perform experiments in three environments: task model, main memory database model and disk resident database model. Our new results show that virtually assigning the priorities of transactions, depending on their behavior, gives a substantial improvement in the number of transactions that meet their deadline in all the three environments.