A Web server typically handles multiple client requests at the same time. These requests must share access to the processor,disk,memory,and network interface at the server. In this section, we discuss techniques for allocating system resources among competing client requests. An euent-driuen server has a single process that alternates between servicing different requests, whereas a processdriuen server allocates request to a separate process.Hybrid schemes attempt to capture the salient features of these two approaches. Although these approaches have been investigated in a variety of contexts before the advent of the Web, the handling of HTTP requests has important implications on the performance trade-offs.