Browse Prior Art Database

Autonomic Session Tuning and Missue Detection

IP.com Disclosure Number: IPCOM000124900D
Original Publication Date: 2005-May-11
Included in the Prior Art Database: 2005-May-11
Document File: 2 page(s) / 18K

Publishing Venue

IBM

Abstract

Autonomic detection of session misuse by J2EE applications.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 55% of the total text.

Page 1 of 2

Autonomic Session Tuning and Missue Detection

Sessions are being used more and more as key state structures in web applications. Session size is a key determinant of application server performance, but it is not easy to see how big sessions are. As a result, it is VERY common for developers to overuse the session, throwing into it everything they might ever need without realizing the performance impact. It is not uncommon to see sessions one or two orders of magnitude larger than the recommended size.

Known solutions & their problems:

It is hard to measure the size of sessions, because in Java, the only way to measure the size of an object is to serialize it, and this is too expensive to do in a production monitoring tool.

Code analysis tools and coding practices can be used to analyze code to detect session misuse but these tools are generally not used in a product environment because they require access to the source code, and because they can be quite expensive to run. Many organizations have established coding practices intended to limit session size, but they rarely seem to work in practice.

Current tooling requires the application server customer to be aware that they may have a problem and be very familiar with their application and sessions impact on performance (i.e be very skilled).

Session problems often only come up in production and current solutions simply track sessions but can not do anything about the problem while it is being solved.

How it work...