Network Deployment (Distributed operating systems), v8.0 > Develop and deploying applications > Develop OSGi applications > Sample OSGi applications


OSGi blog sample application

The blog sample application is a traditional blogging application, used for publishing essay-length articles and allowing readers to comment on them. This sample application shows how to write and package bundles into an enterprise bundle archive (EBA) file. The sample includes example code for Blueprint management, bean injection, using and publishing services from and to the OSGi service registry, and the use of Java persistence.


Before you begin

The blog sample application consists of five bundles, but can be run with four bundles because the fifth bundle is an upgrade to the persistence bundle.

The sample application requires that the supplied com.ibm.samples.websphere.osgi.logging.api.jar and com.ibm.samples.websphere.osgi.logging.impl.jar bundles are installed into the internal bundle repository. The following procedure describes how to do this. These two JAR files, and the blog sample EBA file, are provided in the installableApps directory of the OSGi_blogSample.zip compressed archive file.


About this task

The bundles are divided into the following functional areas:

We can use scripts to completely install the sample application, or you can use scripts to complete the initial configuration of the application then use the administrative console to install the application into the application server. We can also use, modify and remove the sample, and upgrade the persistence service provided by the sample.


Procedure

Configure and install the blog sample.

We can either configure and install the sample using scripts (the first optional step below), or you can configure the sample using scripts then install the sample using the administrative console (the second optional step below).

The scripts that you use to do this are provided in the scripts directory of the OSGi_blogSample.zip compressed archive file. The createBlogDb.sql script contains the necessary configuration commands to create the required Derby database and associated tables. The blogSampleInstall.py script contains the required jython to create data sources, and to install the blog sample with default configuration. You should fully qualify the path to the script if you do not run it from the directory that contains the script.

If we have data sources already defined in the environment, and these data sources have the same name as the data sources specified in the blogSampleInstall.py script, then the sample might not install and run successfully.

In the following steps substitute your own values for the variables WAS_HOME, profileName, serverName, nodeName, blogSample.eba_Location, com.ibm.samples.websphere.osgi.logging.api.jar_Location, com.ibm.samples.websphere.osgi.logging.impl.jar_Location, uncompressed_sample_directory, and path_to_ant.

Use the blog sample.

Upgrade the persistence service.

To add the new persistence bundle to the blog application, use the administrative console to add the com.ibm.samples.websphere.osgi.blog.persistence_1.1.0.jar file to the internal bundle repository, then modify the asset to expect a new bundle version.

Modify the blog sample

All the source code for this application is provided in sub-directories below the uncompressed_sample_dir directory. Each part of the project has its own ant build.xml script. To build the whole application into a newly-available EBA file, you use the ant build.xml file located in the uncompressed_sample_dir directory. WAS ships a version of ant in its bin directory called ws_ant.

To build the sample, you also need several JAR files on the ant classpath.

To simplify matters, you can edit the was.root property in the supplied build.properties file to point to these JAR files.

Remove the blog sample.

To remove the blog sample, you complete the following 3 steps:

  1. Remove the application configuration and the data sources, either by using a script (the first optional step below), or by using the administrative console (the second optional step below).
  2. Remove the shared logging bundles from the internal bundle repository.
  3. Remove the database.

Only remove the bundles if we have no other applications installed that use them. Both the blog and blabber sample applications use the shared logging bundles, so if both applications are installed these bundles should not be removed.

The uninstall script is provided in the scripts directory of the OSGi_blogSample.zip compressed archive file. The blogSampleUninstall.py script contains the required jython to remove data sources and to remove the installation of the blog sample with default configuration. You should fully qualify the path to the script if you do not run it from the directory that contains the script.

In the following steps substitute your own values for the variables WAS_HOME, and profileName.


Results

The following illustration shows the correctly running sample:

Parent topic: Sample OSGi applications

Related concepts:

About OSGi Applications

Related tasks:

Blabber sample application
Secure OSGi Applications

Related reference:

OSGi Applications: Troubleshooting tips
Sample OSGi applications

Related information:

About OSGi Applications
Develop and deploying an OSGi application
Deploy an OSGi application as a business-level application
Administer bundles in the internal bundle repository
Accept OSGi applications with the Blueprint Container specification
Best practices for developing and working with OSGi applications
Develop enterprise OSGi applications for WAS
Task topic Feedback
Copyright IBM Corporation 2009, 2011. All Rights Reserved.
This information center is powered by Eclipse technology.

+

Search Tips   |   Advanced Search