Modeling server usage for online ticket sales

This article describes the use of a discrete event simulation model to estimate the server power required for the online sale of concert tickets to a required service standard. Data are available on the number of purchases made per hour and the percentage of tickets booked online for previous concerts and we describe how these are used to estimate the number of users in the system. We use bootstrapping to allow us to take account of the variability in this estimate when calculating the confidence intervals for the simulation model outputs. A queuing model is also introduced, which is useful to provide a quick calculation of how busy the server is before running the more computationally-intensive simulation model. A numerical example is used to describe the model and the methodology.