|Home > Documentation > User's guide > Sample Applications|
The recommended practice for Geronimo is to set up a workflow using geronimo plugins and maven from development through test and production. To demonstrate this, the samples are built into plugins suitable for the jetty and tomcat web containers in geronimo. Many samples require database access. This is encapsulated into a separate plugin (sample-datasource).
TODO: Demonstrate how to swap databases using plugins; see also the roller plugin
In any case, to install samples as plugins you need to build them first.
For installation you need a geronimo server with a web container installed, such as the geronimo-jetty-minimal or geronimo-jetty6-javaee5 servers (or the equivalent tomcat variations).
If you are using one of the javaee5 server assemblies you can use the admin console to install the plugins. Go to the plugins page, push the "Show Plugins" button, find the sample(s) you want to install, and click install. Be sure to select only plugins for the web container you have installed (e.g. jetty plugins for the jetty web container). Any necessary dependencies such as the sample-datasource plugin will be installed automatically. There are equivalent command line functions (described next under Installing through gshell) if you are using one of the minimal assemblies that does not include the admin console.
For this you can use any geronimo server such as one of the minimal assemblies or even the framework assembly. After starting geronimo, run ./bin/gsh deploy/list-plugins in another terminal window and follow the instructions. Installing a sample plugin will install the sample-datasource plugin and all other dependencies. Alternatively if you build using
you can find a server with the sample installed in directories such as
-Pit is currently only available in the samples trunk (2.2-SNAPSHOT) and not branches/2.1 (2.1.3-SNAPSHOT) or tags/samples-parent-2.1.2 (2.1.2)
The first time building samples it is required that you build from the top level directory. After the initial build, you can do subsequent builds of specific samples as necessary.
You need svn and maven installed on your system.
Maven v 2.0.9 or later is required for building Geronimo 2.1 samples.
To check out the 2.1.2 samples:
Alternatively, you could choose to work with the active 2.1 branch of samples. This will allow you to work with the latest changes in samples but will also expose you to a branch that is under active development and so things may be in flux from day to day:
svn co https://svn.apache.org/repos/asf/geronimo/samples/branches/2.1 gsamples-2.1
Depending on when you look you may find earlier or released versions of the samples under samples/branches and samples/tags. You can browse in any web browser to see what's there. The trunk for samples always contains samples for the highest level version under development https://svn.apache.org/repos/asf/geronimo/samples/trunk
If you run into the "Could not scan module for TLD files...Filename too long" problem when using Windows platform, please check out samples to a short directory (14 characters or less).
There are minimal integration tests to assure that the sample plugins can be installed on the framework server. To run this tests use
The integration tests are not available prior to Geronimo Samples 2.2
After building the sample projects you can find suitable geronimo plans for a sample <sample> in
(or the equivalent -tomcat directory). For most samples you will need to install the sample-datasource plugin first. Samples can be deployed from the admin console "deploy new" on a suitable server or using gshell ./bin/gsh deploy/deploy. Note that in this case you need to start with a server with all the required bits already installed, not the framework server. Since this installation method is not tested automatically it is more likely to be broken than the plugin method.
You can also install the sample datasource as a javaee artifact using the plan we generate when building the plugin, install the sample datasource using the admin console using the database wizard, and run the sql by hand rather than relying on the geronimo DBInitializationGBean.
Before you install any sample applications, deploy the org.tranql/tranql-connector-derby-embed-xa/1.4/rar connector located in the geronimo repo at org/tranql/tranql-connector-derby-embed-xa/1.4/tranql-connector-derby-embed-xa-1.4.rar with the plan at sample-datasource/target/resources/META-INF/plan.xml using the admin console "deploy new" or gshell. This will have exactly the same effect as installing the sample-datasource plugin.
Create a new datasource using the admin console database wizard, selecting the derby-embed-xa type and naming the database "SampleDatabase" and the datasource "SampleTxDatasource". The admin console will come up with a module Id different from what the samples need, so you have to map them in var/config/artifact_aliases.properties. Insert a line like this:
replacing 2.2-SNAPSHOT with the appropriate sample version and adjusting the right hand side as necessary. Be sure geronimo is stopped when you edit artifact_aliases.properties.
Samples that need database connectivity have the database initialized by default by using a DBInitializationGBean to run an sql script. Alternatively here we describe how to initialize the database by hand. In a production scenario most likely your db will be set up by a DBA rather than either of these methods.
There is a maven generated site for samples which can be used to view javadoc and source. It can be found here:
To navigate to javadoc and source code (xref) for each sample do the following: