Browse Prior Art Database

A method for detecting and identifying performance bottleneck against system integration

IP.com Disclosure Number: IPCOM000212355D
Publication Date: 2011-Nov-08
Document File: 5 page(s) / 106K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a system and methods about how to effectively detect and identify the system performance bottleneck. Currently, Large scaled enterprise’s increasing complex system integration , such as business process system,business audit system,business analysis system, for collaboration has been witnessed everywhere, however, it is very tough for the administrators or users to identify which specific system is causing the performance bottleneck.Latency and utility are being used to monitor each separate system. And this method will pop up warning messages noticing which specific system have exceptions when the real latency and utility surpass the expected values.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 5

A method for detecting and identifying performance bottleneck against system integration

Disclosed is a system and methods about how to effectively detect and identify the system performance bottleneck. Currently, Large scaled enterprise's increasing complex system integration , such as business process system,business audit system,business analysis system. for collaboration has been witnessed everywhere, however, it is very tough for the administrators or users to identify which specific system is causing the performance bottleneck .The common solution is to investigate every system from the very beginning to the end after the performance bottleneck occurred. Why not have a timely and smart way with little cost to inform the administrators or users once exception occurs? Latency and utility are being used to monitor each separate system. Besides this, the total latency is also being used to monitor all the system response time. .And this method will pop up warning messages noticing which specific system have exceptions when the real latency and utility surpass the expected values. Take figure 1 for example.

1


Page 02 of 5

Figure 1

2


Page 03 of 5

My invention is listed below: the total requesting and responding circle is divided into a series of separate phase according to the different system functionality against different product and system allocation recourses

In each detailed phase, the performance signal is monitored by the following both Formula 1 and Formula 2 when this particular system is monitored with the consideration of the combination of latency and utility as a whole

Two measurement factors, latency and utility are being illustrated below in detail:

Firstly, "latency" indicates the responding time to deal with one request circle, that is, the total time includes the real execution time and the additional time for preparing the execution when the operating system assigns the "process" resources and/or wait-in-line queue.

With the increasing load, latency will be impacted when the system resources are allocated to meet this request.

Secondly, "utility" means the percentage of the system resource being used. For example, CPU utility and memory utility are being considered in the measurement factors.

The combination of "latency" and "utility" will be considered as a whole signal to narrow down bottleneck. "Utility" includes CPU utility, memory utility and so on. When the latency is the same as before but CPU utility increases...