Browse Prior Art Database

Characterising the cost of processing messages in a message processing system

IP.com Disclosure Number: IPCOM000015642D
Original Publication Date: 2002-Feb-15
Included in the Prior Art Database: 2003-Jun-20
Document File: 2 page(s) / 42K

Publishing Venue

IBM

Abstract

An algorithm is disclosed to characterise the cost of processing messages (such as ESQL messages) in a message processing system. The intent is to minimise the amount of effort and measurement required to characterise the processing cost.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Characterising the cost of processing messages in a message processing system

An algorithm is disclosed to characterise the cost of processing messages (such as ESQL messages) in a message processing system. The intent is to minimise the amount of effort and measurement required to characterise the processing cost.

    With a product such as the Websphere* MQ Integrator message broker it is possible to vary the level of complexity of processing on a message from setting a single value to performing complex cross checking, database archival or retrieval etc. Such processing cannot be easily characterised without taking many measurements as the complexity of processing and volume of data varies and one has to measure all possibilities. There is a need to be able to characterise message processing with a minimum number of measurements. Otherwise the volume of measurements required becomes unwieldy. Once the costs of using various functions has been identified they can then be used in a tool to provide estimates for message throughput.

    In order to characterise such message processing there are two main factors. These are the size of data being manipulated (the proportion of data) and the complexity of the manipulation. For a message of a given size we may only change a certain proportion of it, so we cannot work on the basis of message size alone. The factors of message proportion and complexity are used to form a two dimensional matrix. This matrix covers a range of 5 possible levels of complexity and 5 different proportions of the data, giving a total of 25 possible entries. We only focus on the extremes of the matrix, marked by the points A, B, C and D. A third dimension is added to the matrix by repeating A,B,C and D for each of a number of different message sizes. A fourth dimension is added by adding persistence levels of the message.

Proportion A B of message
changed C D

Complexity of statement

    The lines A,C and B,D represent the extremes of complexity of the statement which is being executed. A simple statement (line A,C) bein...