Measuring Data Terminal Response Time
Original Publication Date: 1985-Mar-01
Included in the Prior Art Database: 2005-Feb-18
This article concerns the combination of free-running timer hardware 8 (Fig. 1) and the associated microcode used to measure multiple terminal end user response times concurrently, and the storage of this data in a "bar chart" series (Fig. 2) kept for each terminal. Fig. 3 shows usage in a typical system. The average response time for a given terminal can be calculated by dividing the total time (40, Fig. 2) for that terminal by the total transaction count (42, Fig. 2) for that terminal. This monitor system employs a split between microcode and hardware, allowing a simple stable multivibrator and divider circuit to measure the response times of many terminals. The timer 8 (Fig. 1) includes an accurate, free-running oscillator 10 that outputs a pulse every twenty-five milliseconds.