Building Your Own Deployment Solution - Java Deployment with JNLP and WebStart - page 216

T ABLE 7.1 Deploylet Lifecycle Implementation
Model Phase Deploylet Method Activation Sequence

Development — —

Publication —

publish

DS Management — —

AH Install — —

AH Check-Out c → s

resolve

Application Update c → s

applicationUpdate

Application Installation rep → s → c

applicationInstallation

Application Un-Installation c → s

applicationUninstallation

AH Other Management — —

AH Update — —

AH And Application Configuration c → s

applicationConfiguration

Resource Installation c → s

resourceInstallation

JRE Preparation c → s

jrePreparation

On-The-Run Update (see runtime services) c → s

Debugging (see runtime services) c → s

Other On-The-Run Management (see runtime services) c → s

Application Execution initApplication/ startApplication/

StopApplication c → s

The third column indicates the practical calls sequence performed in the Java classes imple-
mentation. By means of such method invocations, client and server sides of the same deploylet
are kept consistent. “c” is the ClientDeploylet instance, and “s” is the ServerDeploylet
counterpart. “rep” is the ServerDeploylet repository. For example, in the Application
Installation (seventh) row of Table 7.1, the lifecycle-related applicationInstallation method
is invoked by the ServerDeploylet repository (implemented in the ServerRepositoryImpl
class) on the newly created ServerDeploylet instance. It will, in turn, invoke its
ClientDeploylet instance respective method. All such details will be clearer when we will
examine the source code.

Table 7.1 documents a practical implementation of a deployment lifecycle in Java. Similar to
an Applet instance, when its init method is invoked, the code contained in the method is
meant to perform the initialization, so each of these methods provides “hooks” for lifecycle-
related actions. For example, by overriding the ServerDeploylet 's stopApplication method,