The response-time distribution in a real-time database with optimistic concurrency control and constant execution times

For a real-time shared-memory database with optimistic concurrency control, an approximation for the transaction response-time distribution is obtained. The model assumes that transactions arrive at the database according to a Poisson process, that every transaction uses an equal number of data-items uniformly chosen, and that the multiprogramming level is bounded. The execution time of all transactions is constant. The behavior of the system is approximated by an M/D/c queue with feedback. The probability that a transaction must be fed back for a rerun depends on the number of transactions that has committed during its execution. Numerical experiments, which compare the approximative analysis with a simulation of the database, show that the approximation of the response-time distribution is quite accurate, even for high system loads.