A Literary Gem on How to Lock in Multi-Threaded Programs

Thread-safe programming, especially when we want to make our application optimized and scalable, is very hard. Michael Suess, in his blog named "Thinking Parallel",

• 10 Ways to Reduce Lock Contentions in Threaded Programs, Michael Suess, Blog "Thinking Parallel", July 31, 2007
  

He refers to the book "Java Concurrency in Practice" for some of these tips, but basically the list is his original.

One feasible and well-known solution is to never write them by ourselves but to let libraries handle it. Doug Lea's book on design patterns in Java thread programming covers most of important techniques for the problem domain and they are now available as the standard java.util.concurrent package in Java.

However, as you know, it may not solve all you need. Then let's get back to the principals Michael has just extracted for us.

Axis2 Execution Framework

At developer.com, Deepal Jayasinghe explains the execution framework of Apache Axis2.
The handler architecture with phases, phase rules and module engagement is well explained in the article.

• Axis2 Execution Framework, developer.com, by Deepal Jayasinghe

This article is worth checking out in addition to the official documentation:

• Axis2 Architecture Guide
• Writing Your Own Axis2 Module
  

Today's Paper: Compiler Optimizations for Nondeferred Reference-Counting Garbage Collection (ISMM 2006)

Compiler Optimizations for Nondeferred Reference-Counting Garbage Collection

Author:
Pramod G. Joisha

ISMM'06
Proceedings of the 5th International Symposium on Memory Management, ISMM 2006, Ottawa, Ontario, Canada, June 10-11, 2006, pp.150-161

http://doi.acm.org/10.1145/1133956.1133976

Apache Tomcat Performance

Frank Ziglar reports on the peformance of Tomcat, one of the most used Sevlet engines in the world (Some statistics can be found here), on Window and Linux platforms.

• Comparing Apache Tomcat Performance Across Platforms, Sys-Con LiNUX, By Frank Ziglar, Apr 28, 2006
  — They have measured performance information to distinguish the differences between the Windows and Linux platforms. Given comparable hardware we found the performance differences almost trivial.

The original (full version?) report, devided into two parts, is available at the website of Web performance, Inc.

• Part 1: Performance and Distinct Error Handling under Memory Load
• Part 2: Performance and Distinct Error Handling under Computational Load

Today's Paper: A Smart Hill-Climbing Algorithm for Application Server Configuration (WWW 2004)

A smart hill-climbing algorithm for application server configuration

Authors:
Bowei Xi, Zhen Liu, Mukund Raghavachari, Cathy H. Xia, Li Zhang

WWW 2004
pp.287-296

http://portal.acm.org/citation.cfm?doid=988672.988711