Browse Prior Art Database

Efficient and dynamically extensible classpath management for Java Application Servers.

IP.com Disclosure Number: IPCOM000013876D
Original Publication Date: 2000-Dec-01
Included in the Prior Art Database: 2003-Jun-18
Document File: 2 page(s) / 30K

Publishing Venue

IBM

Abstract

A Java* Application Server supporting servlets and Java server pages or an Enterprise Java Beans* (EJB) server provides the facility to deploy new applications into a running system. There may be a large number of applications installed on any one system. This creates a large number of archive files (known as Jar files) containing Java class files for the installed applications .

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 61% of the total text.

Page 1 of 2

  Efficient and dynamically extensible classpath management for Java Application Servers.

A Java* Application Server supporting servlets and Java server
pages or an Enterprise Java Beans* (EJB) server provides the
facility to deploy new applications into a running system. There
may be a large number of applications installed on any one
system. This creates a large number of archive files (known as
Jar files) containing Java class files for the installed
applications.

     When the application server receives a client request it
needs to load the Java classes required to service that request.
It does this by searching the Jars in its classpath. In a simple
implementation, every Jar file deployed in the system is added to
the classpath. This makes searching for classes within those Jars
at runtime a lengthy operation since there are many potential
Jars to be examined. It is also desirable for a Java Virtual
Machine* (JVM) within the application server to be able to
dynamically pick up newly installed Jar files without
interrupting application service. The requirement therefore is to
keep the classpath as small as possible for fast loading of
classes, retaining the ability to dynamically extend it without
bringing down the server or restarting any running JVMs the
server may be using.

     The problem is solved by keeping an in-memory (persistent if
required) hash-table independent of any JVM, called the "Jar
Map." When a new Jar file is deployed in the running system the
server adds an entry into t...