Custom Application Architecture
You can program the Custom Application to generate network and server loads according to specified patterns. The Custom Application has the following characteristics:
- More than two hosts can be involved in data transfers and processing. You can specify the sequence of hosts involved in each task. The sequence can begin with a workstation or server, end with a workstation or server, and can use any collection of workstations or servers in between.
- Client-Server interaction is task-based. A task consists of many interactions between a client and a server or between successive servers.
The basic task consists of many “phases”. Each phase consists of a data transfer and/or a processing event, which can occur at any end device. This end device becomes a tier for the application. Subsequent phases are typically set up to occur in a chain, where the destination of one data transfer phase becomes the source of the next data transfer phase. The entire task is complete only when the last phase of the task has been completed. A task need not terminate at the originating node, though this is a common case. Wherever the task completes, a “Response Time” statistic is measured to record the time separating the initiation of the task (e.g., when a simulated user presses the “OK” button) to its completion (e.g., when the “hourglass” disappears).