Generate deployment code from the command line

The ejbdeploy tool performs these tasks:

1. Imports code from the input JAR or EAR file
2. Generates deployment code
3. Runs RMIC
4. Compiles the deployment code
5. Exports code to the output JAR or EAR file

Note that for CMP entity beans, a data definition language (DDL) that can be used to create corresponding database tables for a JAR file is contained within the META-INF directory and entitled Table.ddl.

Code validation

When the EJB validator generates an error, warning, or information message, the location of the source of the message is displayed in the following format (an error message is used as an example):

[Error:] JAR_name(location): message_text

where JAR_name is the name of the archive file, location identifies the source of the message, and message_text is the text of the message.

The location is either a number or text. By default, the value is a line number if a line number is available as a result of .java source files residing in the archive. For example, if the line number is 101, the message would appear as follows:

[Error:] JAR_name(101): message_text

If no line number is available, a text description of the location is provided. A message on a method, for example, would be displayed like this:

[Error:] JAR_name(Method: fully_qualified_method_name, Class:
  fully_qualified_name_of_the_class_that_contains_the_method): message_text

Migration

If you want to migrate an EJB 1.0 JAR to WAS and have modified the exisiting generated deployment code to get it to work with a specific database vendor (for example, changing the case of column names to mixed case), that change is not preserved when you redeploy the JAR file with the ejbdeploy tool.

If you originally used VisualAge for Java to specify a mapping and generate the deployment code, then export the EJB project from VisualAge for Java as an EJB 1.1 JAR. Exporting the file as an EJB 1.1 JAR preserves mapping metadata and the case of table and column names.