Skip Headers
Oracle® Fusion Middleware Getting Started With JAX-WS Web Services for Oracle WebLogic Server
11g Release 1 (10.3.6)

Part Number E13758-06
Go to Documentation Home
Home
Go to Book List
Book List
Go to Table of Contents
Contents
Go to Master Index
Master Index
Go to Feedback page
Contact Us

Go to previous page
Previous
Go to next page
Next
PDF · Mobi · ePub

6 Invoking Web Services

This chapter describes how to invoke a WebLogic Web service using Java API for XML-based Web services (JAX-WS).

This chapter includes the following topics:

For more information about:

Note:

It is assumed in this chapter that, when you invoke a Web service using the client-side artifacts generated by the clientgen or wsdlc Ant tasks, you have the entire set of WebLogic Server classes in your CLASSPATH.

Overview of Web Services Invocation

Invoking a Web service refers to the actions that a client application performs to use the Web service.

There are two types of client applications:

You can invoke a Web service from any Java SE or Java EE application running on WebLogic Server (with access to the WebLogic Server classpath). Support for stand-alone Java applications that are running in an environment where WebLogic Server libraries are not available is not available in this release of JAX-WS.

The sections that follow describe how to use Oracle's implementation of the JAX-WS specification to invoke a Web service from a Java client application. You can use this implementation to invoke Web services running on any application server, both WebLogic and non-WebLogic.

This chapter focuses on how to generate a static Java class of the Service interface implementation for the particular Web service you want to invoke. For information about generating dynamic proxy clients, see "Creating Dynamic Proxy Clients" in Programming Advanced Features of JAX-WS Web Services for Oracle WebLogic Server.

WebLogic Server includes examples of creating and invoking WebLogic Web services in the WL_HOME/samples/server/examples/src/examples/webservices directory, where WL_HOME refers to the main WebLogic Server directory. For detailed instructions on how to build and run the examples, open the WL_HOME/samples/server/docs/index.html Web page in your browser and expand the WebLogic Server Examples->Examples->API->Web Services node.

In addition to the command-line tools described in this section, you can use an IDE such as Oracle JDeveloper or Oracle Enterprise Pack for Eclipse (OEPE) for Web service proxy generation and testing. For more information, see "Using Oracle IDEs to Build Web Services" in Introducing WebLogic Web Services for Oracle WebLogic Server.

Invoking a Web Service from a Java SE Client

Note:

You can invoke a Web service from any Java SE or Java EE application running on WebLogic Server (with access to the WebLogic Server classpath). Invoking a Web service from stand-alone Java applications that are running in an environment where WebLogic Server libraries are not available is not supported in this release for JAX-WS Web services.

The following table summarizes the main steps to create a Java SE application that invokes a Web service.

Note:

It is assumed that you use Ant in your development environment to build your client application, compile Java files, and so on, and that you have an existing build.xml file that you want to update with Web services client tasks. For general information about using Ant in your development environment, see Creating the Basic Ant build.xml File. For a full example of a build.xml file used in this section, see Sample Ant Build File for a Java Client.

Table 6-1 Steps to Invoke a Web Service from a Java SE Client

#
Step Description

1

Set up the environment.

Open a command window and execute the setDomainEnv.cmd (Windows) or setDomainEnv.sh (UNIX) command, located in the bin subdirectory of your domain directory. The default location of WebLogic Server domains is MW_HOME/user_projects/domains/domainName, where MW_HOME is the top-level installation directory of the Oracle products and domainName is the name of your domain.

2

Update your build.xml file to execute the clientgen Ant task to generate the needed client-side artifacts to invoke a Web service.

See Using the clientgen Ant Task To Generate Client Artifacts.

3

Get information about the Web service, such as the signature of its operations and the name of the ports.

See Getting Information About a Web Service.

4

Write the client application Java code that includes code for invoking the Web service operation.

See Writing the Java Client Application Code to Invoke a Web Service.

5

Create a basic Ant build file, build.xml.

See Creating the Basic Ant build.xml File.

6

Compile and run your Java client application.

See Compiling and Running the Client Application.


Using the clientgen Ant Task To Generate Client Artifacts

The clientgen WebLogic Web services Ant task generates, from an existing WSDL file, the client artifacts that client applications use to invoke both WebLogic and non-WebLogic Web services. These artifacts include:

  • The Java class for the Service interface implementation for the particular Web service you want to invoke.

  • JAXB data binding artifacts.

  • The Java class for any user-defined XML Schema data types included in the WSDL file.

For additional information about the clientgen Ant task, such as all the available attributes, see "Ant Task Reference" in the WebLogic Web Services Reference for Oracle WebLogic Server.

Update your build.xml file, adding a call to the clientgen Ant task, as shown in the following example:

<taskdef name="clientgen"
     classname="weblogic.wsee.tools.anttasks.ClientGenTask" />
  <target name="build-client">
     <clientgen
       wsdl="http://${wls.hostname}:${wls.port}/complex/ComplexService?WSDL"
       destDir="clientclasses"
       packageName="examples.webservices.simple_client"
       type="JAXWS"/>
  </target>

Before you can execute the clientgen WebLogic Web service Ant task, you must specify its full Java classname using the standard taskdef Ant task.

You must include the wsdl and destDir attributes of the clientgen Ant task to specify the WSDL file from which you want to create client-side artifacts and the directory into which these artifacts should be generated. The packageName attribute is optional; if you do not specify it, the clientgen task uses a package name based on the targetNamespace of the WSDL. The type is required in this example; otherwise, it defaults to JAXRPC.

In this example, the package name is set to the same package name as the client application, examples.webservices.simple_client. If you set the package name to one that is different from the client application, you would need to import the appropriate class files. For example, if you defined the package name as examples.webservices.complex, you would need to import the following class files in the client application:

import examples.webservices.complex.BasicStruct;
import examples.webservices.complex.ComplexPortType;
import examples.webservices.complex.ComplexService;

Note:

The clientgen Ant task also provides the destFile attribute if you want the Ant task to automatically compile the generated Java code and package all artifacts into a JAR file. For details and an example, see "clientgen" in the WebLogic Web Services Reference for Oracle WebLogic Server.

If the WSDL file specifies that user-defined data types are used as input parameters or return values of Web service operations, clientgen automatically generates a JavaBean class that is the Java representation of the XML Schema data type defined in the WSDL. The JavaBean classes are generated into the destDir directory.

For a full sample build.xml file that contains additional targets from those described in this procedure, such as clean, see Sample Ant Build File for a Java Client.

To execute the clientgen Ant task, along with the other supporting Ant tasks, specify the build-client target at the command line:

prompt> ant build-client

See the clientclasses directory to view the files and artifacts generated by the clientgen Ant task.

Getting Information About a Web Service

You need to know the name of the Web service and the signature of its operations before you write your Java client application code to invoke an operation. There are a variety of ways to find this information.

The best way to get this information is to use the clientgen Ant task to generate the Web service-specific Service files and look at the generated *.java files. These files are generated into the directory specified by the destDir attribute, with subdirectories corresponding to either the value of the packageName attribute, or, if this attribute is not specified, to a package based on the targetNamespace of the WSDL.

  • The ServiceName.java source file contains the getPortName() methods for getting the Web service port, where ServiceName refers to the name of the Web service and PortName refers to the name of the port. If the Web service was implemented with a JWS file, the name of the Web service is the value of the serviceName attribute of the @WebService JWS annotation and the name of the port is the value of the portName attribute of the <WLHttpTransport> child element of the <jws> element of the jwsc Ant task.

  • The PortType.java file contains the method signatures that correspond to the public operations of the Web service, where PortType refers to the port type of the Web service. If the Web service was implemented with a JWS file, the port type is the value of the name attribute of the @WebService JWS annotation.

You can also examine the actual WSDL of the Web service; see Browsing to the WSDL of the Web Service for details about the WSDL of a deployed WebLogic Web service. The name of the Web service is contained in the <service> element, as shown in the following excerpt of the TraderService WSDL:

<service name="TraderService">
    <port name="TraderServicePort"
         binding="tns:TraderServiceSoapBinding">
  ...
    </port>
  </service>

The operations defined for this Web service are listed under the corresponding <binding> element. For example, the following WSDL excerpt shows that the TraderService Web service has two operations, buy and sell (for clarity, only relevant parts of the WSDL are shown):

<binding name="TraderServiceSoapBinding" ...>
    ...
    <operation name="sell">
    ...
    </operation>
    <operation name="buy">
    </operation>
  </binding>

Writing the Java Client Application Code to Invoke a Web Service

In the following code example, a Java application invokes a Web service operation. The application uses standard JAX-WS API code and the Web service-specific implementation of the Service interface, generated by clientgen, to invoke an operation of the Web service.

The example also shows how to invoke an operation that has a user-defined data type (examples.webservices.simple_client.BasicStruct) as an input parameter and return value. The clientgen Ant task automatically generates the Java code for this user-defined data type.

Because the <clientgen> packageName attribute was set to the same package name as the client application, we are not required to import the <clientgen>-generated files.

package examples.webservices.simple_client;
/**
 * This is a simple Java application that invokes the
 * the echoComplexType operation of the ComplexService Web service.
 */
public class Main {
  public static void main(String[] args) { 
    ComplexService test = new ComplexService(); 
    ComplexPortType port = test.getComplexPortTypePort();
    BasicStruct in = new BasicStruct();
    in.setIntValue(999);
    in.setStringValue("Hello Struct");
    BasicStruct result = port.echoComplexType(in);
    System.out.println("echoComplexType called. Result: " + result.getIntValue() + ", " + result.getStringValue());
  }
}

In the preceding example:

  • The following code shows how to create a ComplexPortType stub:

    ComplexService test = new ComplexService(), 
    ComplexPortType port = test.getComplexPortTypePort();
    

    The ComplexService class implements the JAX-WS Service interface. The getComplexServicePortTypePort() method is used to return an instance of the ComplexPortType stub implementation.

  • The following code shows how to invoke the echoComplexType operation of the ComplexService Web service:

    BasicStruct result = port.echoComplexType(in);
    

    The echoComplexType operation returns the user-defined data type called BasicStruct.

Compiling and Running the Client Application

Add javac tasks to the build-client target in the build.xml file to compile all the Java files (both of your client application and those generated by clientgen) into class files, as shown by the bold text in the following example:

<target name="build-client">
    <clientgen
      wsdl="http://${wls.hostname}:${wls.port}/complex/ComplexService?WSDL"
      destDir="clientclasses"
      packageName="examples.webservices.simple_client"
      type="JAXWS"/>
    <javac
      srcdir="clientclasses" 
      destdir="clientclasses"
      includes="**/*.java"/>
    <javac
      srcdir="src" 
      destdir="clientclasses"
      includes="examples/webservices/simple_client/*.java"/>
  </target>

In the example, the first javac task compiles the Java files in the clientclasses directory that were generated by clientgen, and the second javac task compiles the Java files in the examples/webservices/simple_client subdirectory of the current directory; where it is assumed your Java client application source is located.

In the preceding example, the clientgen-generated Java source files and the resulting compiled classes end up in the same directory (clientclasses). Although this might be adequate for prototyping, it is often a best practice to keep source code (even generated code) in a different directory from the compiled classes. To do this, set the destdir for both javac tasks to a directory different from the srcdir directory. To run the client application, add a run target to the build.xml that includes a call to the java task, as shown below:

<path id="client.class.path">
    <pathelement path="clientclasses"/>
    <pathelement path="${java.class.path}"/>
</path>
<target name="run" >
    <java 
       fork="true" 
       classname="examples.webServices.simple_client.Main"
       failonerror="true" >
       <classpath refid="client.class.path"/>
</target>

The path task adds the clientclasses directory to the CLASSPATH. The run target invokes the Main application, passing it the URL of the deployed Web service as its single argument.

See Sample Ant Build File for a Java Client for a full sample build.xml file that contains additional targets from those described in this procedure, such as clean.

Rerun the build-client target to regenerate the artifacts and recompile into classes, then execute the run target to invoke the echoStruct operation:

prompt> ant build-client run

You can use the build-client and run targets in the build.xml file to iteratively update, rebuild, and run the Java client application as part of your development process.

Sample Ant Build File for a Java Client

The following example shows a complete build.xml file for generating and compiling a Java client. See Using the clientgen Ant Task To Generate Client Artifacts and Compiling and Running the Client Application for explanations of the sections in bold.

<project name="webservices-simple_client" default="all">
  <!-- set global properties for this build -->
  <property name="wls.hostname" value="localhost" />
  <property name="wls.port" value="7001" />
  <property name="example-output" value="output" />
  <property name="clientclass-dir" value="${example-output}/clientclass" />
  <path id="client.class.path">
    <pathelement path="${clientclass-dir}"/>
    <pathelement path="${java.class.path}"/>
  </path>
  <taskdef name="clientgen"
    classname="weblogic.wsee.tools.anttasks.ClientGenTask" />
  <target name="clean" >
    <delete dir="${clientclass-dir}"/>
  </target>
  <target name="all" depends="clean,build-client,run" />
  <target name="build-client">
    <clientgen
      wsdl="http://${wls.hostname}:${wls.port}/complex/ComplexService?WSDL"
      destDir="${clientclass-dir}"
      packageName="examples.webservices.simple_client"
      type="JAXWS"/>
    <javac
      srcdir="${clientclass-dir}" destdir="${clientclass-dir}"
      includes="**/*.java"/>
    <javac
      srcdir="src" destdir="${clientclass-dir}"
      includes="examples/webservices/simple_client/*.java"/>
  </target>
  <target name="run" >
    <java fork="true"
          classname="examples.webservices.simple_client.Main"
          failonerror="true" >
      <classpath refid="client.class.path"/>
    </java>
  </target>
</project>

Invoking a Web Service from a WebLogic Web Service

Invoking a Web service from within a WebLogic Web service is similar to invoking one from a Java SE application, as described in Invoking a Web Service from a Java SE Client, with the following variation:

This section describes the differences between invoking a Web service from a client in a Java EE component, specifically another Web service, and invoking from a Java SE client. It is assumed that you have read and understood Invoking a Web Service from a Java SE Client. It is also assumed that you use Ant in your development environment to build your client application, compile Java files, and so on, and that you have an existing build.xml that builds a Web service that you want to update to invoke another Web service.

The following list describes the changes you must make to the build.xml file that builds your client Web service, which will invoke another Web service. See Sample build.xml File for a Web Service Client for the full sample build.xml file:

The following list describes the changes you must make to the JWS file that implements the client Web service; see Sample JWS File That Invokes a Web Service for the full JWS file example.

Sample build.xml File for a Web Service Client

The following sample build.xml file shows how to create a Web service that itself invokes another Web service; the relevant sections that differ from the build.xml for building a simple Web service that does not invoke another Web service are shown in bold.

The build-service target in this case is very similar to a target that builds a simple Web service; the only difference is that the jwsc Ant task that builds the invoking Web service also includes a <clientgen> child element of the <jws> element so that jwsc also generates the required JAX-WS client stubs.

<project name="webservices-service_to_service" default="all">
  <!-- set global properties for this build -->
  <property name="wls.username" value="weblogic" />
  <property name="wls.password" value="weblogic" />
  <property name="wls.hostname" value="localhost" />
  <property name="wls.port" value="7001" />
  <property name="wls.server.name" value="myserver" />
  <property name="ear.deployed.name" value="ClientServiceEar" />
  <property name="example-output" value="output" />
  <property name="ear-dir" value="${example-output}/ClientServiceEar" />
  <property name="clientclass-dir" value="${example-output}/clientclasses" />
  <path id="client.class.path">
    <pathelement path="${clientclass-dir}"/>
    <pathelement path="${java.class.path}"/>
  </path>
  <taskdef name="jwsc"
    classname="weblogic.wsee.tools.anttasks.JwscTask" />
  <taskdef name="clientgen"
    classname="weblogic.wsee.tools.anttasks.ClientGenTask" />
  <taskdef name="wldeploy"
    classname="weblogic.ant.taskdefs.management.WLDeploy"/>
  <target name="all" depends="clean,build-service,deploy,client" />
  <target name="clean" depends="undeploy">
    <delete dir="${example-output}"/>
  </target>
  <target name="build-service">
    <jwsc
        srcdir="src"
        destdir="${ear-dir}" >
        <jws
         file="examples/webservices/service_to_service/ClientServiceImpl.java"
         type="JAXWS">
          <clientgen
                wsdl="http://${wls.hostname}:${wls.port}/complex/ComplexService?WSDL"
                packageName="examples.webservices.complex" />
        </jws>
    </jwsc>
  </target>
  <target name="deploy">
    <wldeploy action="deploy" name="${ear.deployed.name}"
      source="${ear-dir}" user="${wls.username}"
      password="${wls.password}" verbose="true"
      adminurl="t3://${wls.hostname}:${wls.port}"
      targets="${wls.server.name}" />
  </target>
  <target name="undeploy">
    <wldeploy action="undeploy" name="${ear.deployed.name}"
      failonerror="false"
      user="${wls.username}"
      password="${wls.password}" verbose="true"
      adminurl="t3://${wls.hostname}:${wls.port}"
      targets="${wls.server.name}" />
  </target>
  <target name="client">
    <clientgen
      wsdl="http://${wls.hostname}:${wls.port}/ClientService/ClientService?WSDL"
      destDir="${clientclass-dir}"
      packageName="examples.webservices.service_to_service.client"
      type="JAXWS"/>
    <javac
      srcdir="${clientclass-dir}" destdir="${clientclass-dir}"
      includes="**/*.java"/>
    <javac
      srcdir="src" destdir="${clientclass-dir}"
      includes="examples/webservices/service_to_service/client/**/*.java"/>
  </target>
  <target name="run">
    <java classname="examples.webservices.service_to_service.client.Main"
          fork="true"
          failonerror="true" >
          <classpath refid="client.class.path"/>
    </java>
  </target>
</project>

Sample JWS File That Invokes a Web Service

The following sample JWS file, called ClientServiceImpl.java, implements a Web service called ClientService that has an operation that in turn invokes the echoComplexType operation of a Web service called ComplexService. This operation has a user-defined data type (BasicStruct) as both a parameter and a return value. The relevant code is shown in bold and described after the example.

package examples.webservices.service_to_service;

import javax.jws.WebService;
import javax.jws.WebMethod;
import javax.xml.ws.WebServiceRef;

// Import the BasicStruct data type, generated by clientgen and used
// by the ComplexService Web Service
import examples.webservices.complex.BasicStruct;

// Import the JAX-WS stubs generated by clientgen for invoking
// the ComplexService Web service.
import examples.webservices.complex.ComplexPortType;
import examples.webservices.complex.ComplexService;

@WebService(name="ClientPortType", serviceName="ClientService",
            targetNamespace="http://examples.org")
public class ClientServiceImpl {
// Use the @WebServiceRef annotation to define a reference to a Web service.
  @WebServiceRef()
  ComplexService test;

  @WebMethod()
  public String callComplexService(BasicStruct input, String serviceUrl) 
  {
    // Create a port stub to invoke ComplexService
    ComplexPortType port = test.getComplexPortTypePort();

    // Invoke the echoComplexType operation of ComplexService
    BasicStruct result = port.echoComplexType(input);
    System.out.println("Invoked ComplexPortType.echoComplexType." );
    return "Invoke went okay!  Here's the result: '" + result.getIntValue() + 
           ",  " + result.getStringValue() + "'";
  }
}

Follow these guidelines when programming the JWS file that invokes another Web service; code snippets of the guidelines are shown in bold in the preceding example:

  • Import any user-defined data types that are used by the invoked Web service. In this example, the ComplexService uses the BasicStruct JavaBean:

    import examples.webservices.complex.BasicStruct;
    
  • Import the JAX-WS interfaces of the ComplexService Web service; the stubs are generated by the <cliengen> child element of <jws>:

    import examples.webservices.complex.ComplexPortType;
    import examples.webservices.complex.ComplexService;
    
    
  • Define a reference to a Web service and an injection target for it using the @WebServiceRef annotation:

    @WebServiceRef()
    ComplexService service;
    

    Alternatively, you can create a proxy stub to the ComplexService Web service, as shown below:

    ComplexService test = new ComplexService();  
    
  • Return an instance of the ComplexPortType stub implementation by calling the getComplexPortTypePort() operation on the Web service reference:

    ComplexPortType port = service.getComplexPortTypePort();
    
  • Invoke the echoComplexType operation of ComplexService using the port you just instantiated:

    BasicStruct result = port.echoComplexType(input);
    

Configuring Web Service Clients

By default, Web service clients use the Web service configuration defined for the server. You can override the configuration settings used by the Web service client using one of the following methods:

Defining a Web Service Reference Using the @WebServiceRef Annotation

The @WebServiceRef annotation enables you to define a reference to a Web service and attach the configuration of the Web service to the client instance.

For example, in the following code excerpt, @WebServiceRef is used to attach the configuration for MyReliableEchoService to the client's Web service instance. The port that is subsequently created and initialized uses the properties defined for MyReliableEchoService service reference in the weblogic.xml for the Web application.

package wsrm_jaxws.example;
import java.xml.ws.WebService;
import java.xml.ws.WebServiceRef;
import wsrm_jaxws.example.client_service.*;
import wsrm_jaxws.example.client_service.EchoResponse;
...
@WebService
public class ClientServiceImpl {

    @WebServiceRef(name="MyServiceRef")
    private ReliableEchoService service;
    private ReliableEchoPortType port = null;

    @PostConstruct
    public void initPort() {
        port = service.getReliableEchoPort();
    ...
    }
}

Example 6-1 shows an example of a weblogic.xml file that contains a Web service reference description. For information about the reliable messaging properties shown in this example, see "Configuring Reliable Messaging" in Programming Advanced Features of JAX-WS Web Services for Oracle WebLogic Server.

Example 6-1 Example weblogic.xml File Containing Web Service Reference Description

<?xml version='1.0' encoding='UTF-8'?>
<weblogic-web-app xmlns="http://xmlns.oracle.com/weblogic/weblogic-web-app">
   <service-reference-description>
     <!-- Any name you want, but use this same name on
          @WebServiceRef(name=<my name>). This anno goes on the service
          field in your client container -->
      <service-ref-name>MyServiceRef</service-ref-name>
      <!-- Use / and any path within the web app to get a local WSDL, or
           use a resource name as defined by the Java ClassLoader, or use an
           absolute/external URL you can guarantee is deployed when this web
           app deploys -->
      <wsdl-url>/WEB-INF/wsdls/ReliableEcho.wsdl</wsdl-url>
      <!-- One or more port-infos, one for each type of port/stub you'll create
           in your JWS -->
      <port-info>
         <!-- The local name of wsdl:port (not portType). The Java type for this
              port, when created from the @WebServiceRef JWS field, will contain,
              in RequestContext, the props you define below -->
         <port-name>ReliableEchoPort</port-name>
 
         <!-- Any prop name/value pairs you want to show up on you service stub
              The Java type for this port, when created from the @WebServiceRef JWS field, 
              will contain, in RequestContext, the stub-props you define below -->
 
         <!-- RM Source Properties -->
 
         <stub-property>
            <name>weblogic.wsee.wsrm.BaseRetransmissionInterval</name>
            <value>PT30S</value>
         </stub-property>
 
         <stub-property>
            <name>weblogic.wsee.wsrm.RetransmissionExponentialBackoff</name>
            <value>true</value>
         </stub-property>
 
         <!-- RM Destination Properties -->
 
          <stub-property>
            <name>weblogic.wsee.wsrm.RetryCount</name>
            <value>5</value>
         </stub-property>
 
         <stub-property>
            <name>weblogic.wsee.wsrm.RetryDelay</name>
            <value>PT30S</value>
         </stub-property>
 
         <stub-property>
            <name>weblogic.wsee.wsrm.AcknowledgementInterval</name>
            <value>PT5S</value>
         </stub-property>
 
         <stub-property>
            <name>weblogic.wsee.wsrm.NonBufferedDestination</name>
            <value>true</value>
         </stub-property>
 
         <!-- RM Source *or* Destination Properties -->
 
         <stub-property>
            <name>weblogic.wsee.wsrm.InactivityTimeout</name>
            <value>PT5M</value>
         </stub-property>
 
         <stub-property>
            <name>weblogic.wsee.wsrm.SequenceExpiration</name>
            <value>PT10M</value>
         </stub-property>
 
   </port-info>
 
   </service-reference-description>
   <wl-dispatch-policy>weblogic.wsee.mdb.DispatchPolicy</wl-dispatch-policy>
</weblogic-web-app>

Managing Client Identity

Web services enable you to assign any meaningful name to a client, which is represented as the client identity (client ID). This client ID is used to group statistics and other monitoring information, and for reporting runtime validations, and so on.

For on-server clients (clients running in a container within a WebLogic Server instance), the client ID can be generated in one of the following ways:

Note:

Although optional, Oracle strongly recommends that you define the client ID explicitly.

The weblogic.wsee.jaxws.persistence.ClientIdentityFeature client feature enables Web service clients to set and access the Web service client ID. The following table summarizes the ClientIdentityFeature methods.

Table 6-2 Methods of ClientIdentityFeature for Setting and Accessing Client ID

Method Description

getClientID()

Gets the currently defined client ID for the Web service port.

setClientID()

Sets the client ID for the Web service port.

In addition, you can set the client ID by passing it as an argument when instantiating the ClientIdentityFeature object. For example:

ClientIdentityFeature clientIDFeature = new 
 ClientIdentityFeature("MyBackendServiceAsyncClient"); 

dispose()

Disposes the client ID.

If a client ID is not disposed of explicitly, it will be done when the container for the client instances that use the client ID is deactivated (for example, the host Web application or EJB is deactivated). For more information, see Client Identity Lifecycle.


The following sections describe the methods for managing the client ID:

Defining the Client ID During Port Initialization

To provide its client ID, the Web service client can pass an instance of the ClientIdentityFeature containing the client ID to the Web service port at initialization time.

The client ID must be unique within the Web application or EJB that contains the client. It is recommended that the client ID appropriately reflect the business purpose. In order to ensure that the client ID is unique, the system prepends the names of the containing server, application, and component (Web application or EJB) to the client ID.

Notes:

Care should be taken when choosing a client ID. If a client instance is created with the same client ID as an existing client instance, the two client instances will be treated as the same instance. No exception will be thrown to alert you to the duplication.

The following example demonstrates this method of specifying the client ID. It is recommended that you close the client instance once all processing has been complete, as shown.

This example is excerpted from "Web Service Client Best Practices Example" in Programming Advanced Features of JAX-WS Web Services for Oracle WebLogic Server.

Example 6-2 Example of Specifying the Client ID During Port Initialization

import javax.servlet.*;
import javax.xml.ws.*;
import weblogic.jws.jaxws.client.ClientIdentityFeature;
. . . 
public class BestPracticeAsyncClient
  extends GenericServlet {
...
  private BackendServiceService _service;
...
    // Client ID
    ClientIdentityFeature clientIdFeature = 
        new ClientIdentityFeature("MyBackendServiceAsyncClient");
    features.add(clientIdFeature);
...
     _features = features.toArray(new WebServiceFeature[features.size()]);
...
    BackendService port = _service.getBackendServicePort(_features);
...
    ((java.io.Closeable)_port).close(); 
  }
}

Accessing the Server-generated Client ID

Note:

As described in this section, in order to ensure that the client ID is unique, the server-generated version may be long and difficult to read. To guarantee that the client ID is presented in a user-friendly format, it is recommended that you define the client ID during port initialization, as described in Defining the Client ID During Port Initialization.

Client IDs that are generated automatically by the server use the following format:

applicationname[_applicationversion]:componentname:uniqueID

Where:

  • applicationname—Name of the application hosting the client.

  • applicationversion—Version of the application. Only used if multiple versions of the same application is running simultaneously.

  • componentname—Name of the component (Web application or EJB) hosting the client.

  • uniqueID—Calculated based on the information that is available when the client instance is created. The uniqueID is constructed by choosing one of the following (whichever is available):

    • Web service reference name, as defined by the @WebServiceRef annotation.

    • [portNamespaceURI:portLocalName][:][endpointAddress]—port name, endpoint address, or both (separated by a colon).

    • Port class simple name.

    The following information, when available, may also be concatenated to the uniqueID, separated by a colon (:), in the order presented below:

    • WSDL location (minus ?wsdl)

    • Features used to create the client instance, represented by the features class name and separated by dash (-).

For example, assume that you deploy a Web service client with the following information associated with it:

  • Application name: example

  • Component: Web application called BestPracticeClient

  • Port name: http://example/BackendServicePort

  • Port class: BackendService

  • WSDL: jar:file:/E:/p4/dev/src1034/wls/modules/wsee/test/server/build/output/example/BackendService.war!/WEB-INF/BackendServiceService.wsdl

The server-generated client ID will be:

example:BestPracticeClient:http://example/:BackendServicePort:jar:file:/E:/p4/dev/src1034/wls/modules/wsee/test/server/build/output/example/BackendService.war!/WEB-INF/BackendServiceService.wsdl:AsyncClientTransportFeature()-ClientIdentityFeature

Each time the code is executed, assuming it is in the same containment hierarchy, the same client ID is generated. This provides a stable client ID that can be used across server VM instances and allows for asynchronous responses to be delivered to the client even after a server restart.

Note:

A given Client ID can be used from multiple locations in the client code, but care should be taken to initialize any port or Dispatch instance that uses that client ID in the same way (same features, service, and so on) as was used in any other location for that client ID.

For best practice information on the recommended approach to client instance (port or Dispatch) initialization, see "Roadmap for Developing Web Service Clients" in Getting Started With JAX-WS Web Services for Oracle WebLogic Server.

The following example demonstrates how to access the server-generated client ID. This example is excerpted from "Web Service Client Best Practices Example" in Programming Advanced Features of JAX-WS Web Services for Oracle WebLogic Server.

Example 6-3 Example of Accessing the Server-generated Client ID

...
    // Create a port without explicitly defining the client ID to view the client ID that is
    // generated automatically.
    ClientIdentityFeature dummyClientIdFeature = new ClientIdentityFeature(null);
    BackendService dummyPort = _service.getBackendServicePort(dummyClientIdFeature);
    System.out.println("Generated Client Identity is: " + dummyClientIdFeature.getClientId());

    // Best Practice: Explicitly close client instances when processing is complete.
    // If not closed, the port will be closed automatically when it goes out of scope.
    // Note, this client ID will remain registered and visible until our
    // container (Web application) is undeployed.
    ((java.io.Closeable)dummyPort).close();

Client Identity Lifecycle

A client ID is registered with the Web services runtime when the first client instance (port or Dispatch instance) using the client ID is created. Any asynchronous response endpoint associated with the client instances is also tracked along with the registered client ID.

The client ID remains registered until one of the following occurs:

  • The client ID is explicitly disposed using the dispose() method on ClientIdentityFeature, as described in Table 6-2.

  • The container for the client instances that use the client ID is deactivated (for example, the host Web application or EJB is deactivated).

Using a Proxy Server When Invoking a Web Service

You can use a proxy server to proxy requests from a client application to an application server (either WebLogic or non-WebLogic) that hosts the invoked Web service. You typically use a proxy server when the application server is behind a firewall. You can specify the proxy server in your client application using Java system properties. There are two ways to specify the proxy server in your client application: programmatically using the WebLogic ClientProxyFeature API or using system properties.

Using the ClientProxyFeature API to Specify the Proxy Server

You can programmatically specify within the Java client application itself the details of the proxy server that will proxy the Web service invoke using the weblogic.wsee.jaxws.proxy.ClientProxyFeature API. For more about the ClientProxyFeature API, see the Oracle WebLogic Server API Reference.

The proxy server settings defined by the ClientProxyFeature override the settings defined at the JVM-level, as described in Using System Properties to Specify the Proxy Server.

Note:

The ClientProxyFeature configures the port for WebLogic HTTP over SSL. It is recommended that you configure SSL for WebLogic Server. For more information, see "Configuring SSL" in Securing Oracle WebLogic Server.

You can configure the proxy server information using the ClientProxyFeature and pass the feature as an argument when creating the Web service port, as shown in the following example.

Example 6-4 Pass ClientProxyFeature as an Argument When Creating Port

package examples.webservices.simple_client;
import weblogic.wsee.jaxws.proxy
public class Main {
  public static void main(String[] args) { 
    ComplexService test = new ComplexService(); 
    ClientProxyFeature cpf = new ClientProxyFeature();
    cpf.setProxyHost("localhost");
    cpf.setProxyPort(8888);
    cpf.setProxyUserName("proxyu");
    cpf.setProxyPassword("proxyp");
    ComplexPortType port = test.getComplexPortTypePort(cpf);
    BasicStruct in = new BasicStruct();
    in.setIntValue(999);
    in.setStringValue("Hello Struct");
    BasicStruct result = port.echoComplexType(in);
    System.out.println("echoComplexType called. Result: " + result.getIntValue() + ", " + result.getStringValue());
  }
}

Alternatively, you can configure the proxy server information after the port is created, as shown in the following example. In this case, you execute the attachsPort() method to attach the ClientProxyFeature to the existing port.

Example 6-5 Configuring the ClientProxyFeature After Creating the Port

package examples.webservices.simple_client;
import weblogic.wsee.jaxws.proxy
public class Main {
  public static void main(String[] args) { 
    ComplexService test = new ComplexService(); 
    ComplexPortType port = test.getComplexPortTypePort();
    ClientProxyFeature cpf = new ClientProxyFeature();
    cpf.setProxyHost("localhost");
    cpf.setProxyPort(8888);
    cpf.setProxyUserName("proxyu");
    cpf.setProxyPassword("proxyp");
    cpf.attachsPort(port);
    BasicStruct in = new BasicStruct();
    in.setIntValue(999);
    in.setStringValue("Hello Struct");
    BasicStruct result = port.echoComplexType(in);
    System.out.println("echoComplexType called. Result: " + result.getIntValue() + ", " + result.getStringValue());
  }
}

If after configuring the ClientProxyFeature and attaching it to the port you want to disable the client proxy settings, you set the proxy port to a negative value. For example:

Example 6-6 Disabling Client Proxy Settings

. . .
    ClientProxyFeature cpf = new ClientProxyFeature();
    cpf.setProxyPort(-1);\
    cpf.attachsPort(port);
. . .

Using System Properties to Specify the Proxy Server

To use system properties to specify the proxy server, write your client application in the standard way, and then specify Java system properties when you execute the client application.

The following table summarizes the Java system properties.

Note:

In this case, the proxySet system property must not be set. If the proxySet system property is set to (proxySet=false), proxy properties will be ignored and no proxy will be used.

Table 6-3 Java System Properties Used to Specify Proxy Server

Property Description

http.proxyHost=proxyHost or https.proxyHost=proxyHost

Name of the host computer on which the proxy server is running. Use https.proxyHost for HTTP over SSL.

http.proxyPort=proxyPort or https.proxy.Port=proxyPort

Port to which the proxy server is listening. Use https.proxyPort for HTTP over SSL.

http.nonProxyHosts=hostname | hostname | ...

List of hosts that should be reached directly, bypassing the proxy. Separate each host name using a | character. This property applies to both HTTP and HTTPS.


The following excerpt from an Ant build script shows an example of setting Java system properties when invoking a client application called clients.InvokeMyService:

<target name="run-client">
     <java fork="true"
           classname="clients.InvokeMyService"
           failonerror="true">
       <classpath refid="client.class.path"/>
       <arg line="${http-endpoint}"/>
       <jvmarg line=
         "-Dhttp.proxyHost=${proxy-host} 
         -Dhttp.proxyPort=${proxy-port}
         -Dhttp.nonProxyHosts=${mydomain}"
       />
     </java>
   </target>

Client Considerations When Redeploying a Web Service

WebLogic Server supports production redeployment, which means that you can deploy a new version of an updated WebLogic Web service alongside an older version of the same Web service.

WebLogic Server automatically manages client connections so that only new client requests are directed to the new version. Clients already connected to the Web service during the redeployment continue to use the older version of the service until they complete their work, at which point WebLogic Server automatically retires the older Web service.

You can continue using the old client application with the new version of the Web service, as long as the following Web service artifacts have not changed in the new version:

If any of these artifacts have changed, you must regenerate the JAX-WS stubs used by the client application by re-running the clientgen Ant task.

For example, if you change the signature of an operation in the new version of the Web service, then the WSDL file that describes the new version of the Web service will also change. In this case, you must regenerate the JAX-WS stubs. If, however, you simply change the implementation of an operation, but do not change its public contract, then you can continue using the existing client application.

Client Considerations When Web Service and Client Are Deployed to the Same Managed Server

If a Web service and client are deployed to the same Managed Server, and one of the following is true:

Then, when you restart the Managed Server on which the Web service and client are deployed, the Web service client may fail to redeploy, regardless of the deployment order, because the applications are deployed initially in administration mode, and later transition to production mode to accept HTTP requests. In this situation, you must restart the application manually once the server has restarted.

If a Web service and client are deployed to the same Managed Server, to avoid this situation, it is recommended that you package the WSDL as part of the Web service application and refer to the packaged version from the @WebServiceRef annotation.