JBoss.orgCommunity Documentation

Drools Integration User Guide

by The JBoss Drools team

Version 6.1.0-SNAPSHOT

1. Apache Camel Integration
1.1. Camel
2. Spring Integration
2.1. Integration with Drools Expert
2.1.1. Define a Resource
2.1.2. Event Listeners
2.2. Integration with jBPM Human Task
2.2.1. How to configure Spring with jBPM Human task
3. Drools Camel Server
3.1. Introduction
3.2. Deployment
3.3. Configuration
3.3.1. REST/Camel Services configuration
4. Drools Commands
4.1. API
4.1.1. XStream
4.1.2. JSON
4.1.3. JAXB
4.2. Commands supported
4.2.1. BatchExecutionCommand
4.2.2. InsertObjectCommand
4.2.3. RetractCommand
4.2.4. ModifyCommand
4.2.5. GetObjectCommand
4.2.6. InsertElementsCommand
4.2.7. FireAllRulesCommand
4.2.8. StartProcessCommand
4.2.9. SignalEventCommand
4.2.10. CompleteWorkItemCommand
4.2.11. AbortWorkItemCommand
4.2.12. QueryCommand
4.2.13. SetGlobalCommand
4.2.14. GetGlobalCommand
4.2.15. GetObjectsCommand
5. OSGi Integration
6. JMX monitoring with RHQ/JON
6.1. Introduction
6.1.1. Enabling JMX monitoring in a Drools application
6.1.2. Installing and running the RHQ/JON plugin
1.1. Camel

Camel provides a light weight bus framework for getting information into and out of Drools.

Drools introduces two elements to make easy integration.

Drools can be configured like any normal camel component, but notice the policy that wraps the drools related segments. This will route all payloads to ksession1

Example 1.1. Drools EndPoint configured with the CXFRS producer


<bean id="droolsPolicy" class="org.drools.camel.component.DroolsPolicy" />



<camelContext id="camel" xmlns="http://camel.apache.org/schema/spring">  

   <route>

      <from uri="cxfrs://bean://rsServer"/>

         <policy ref="droolsPolicy">

            <unmarshal ref="xstream" />       

            <to uri="drools:node1/ksession1" />

            <marshal ref="xstream" />

       </policy>

    </route>

It is possible to not specify the session in the drools endpoint uri, and instead "multiplex" based on an attribute or header. In this example the policy will check either the header field "DroolsLookup" for the named session to execute and if that isn't specified it'll check the "lookup" attribute on the incoming payload. It then attempts to "lookup" the session from the execution-node context and execute against it.

Example 1.2. Drools EndPoint configured with the CXFRS producer


<bean id="droolsPolicy" class="org.drools.camel.component.DroolsPolicy" />



<camelContext id="camel" xmlns="http://camel.apache.org/schema/spring">  

   <route>

      <from uri="cxfrs://bean://rsServer"/>

         <policy ref="droolsPolicy">

            <unmarshal ref="xstream" />       

            <to uri="drools:node1" />

            <marshal ref="xstream" />

       </policy>

    </route>

Example 1.3. Java Code to execute against Route from a Spring and Camel Context

public class MyTest extends CamelSpringTestSupport {
    
    @Override
    protected AbstractXmlApplicationContext createApplicationContext() {        
        return new ClassPathXmlApplicationContext("org/drools/camel/component/CxfRsSpring.xml");
    }  

    public void test1() throws Exception {
        String cmd = "";
        cmd += "<batch-execution lookup=\"ksession1\">\n";
        cmd += "  <insert out-identifier=\"salaboy\">\n";
        cmd += "      <org.drools.pipeline.camel.Person>\n";
        cmd += "         <name>salaboy</name>\n";
        cmd += "      </org.drools.pipeline.camel.Person>\n";
        cmd += "   </insert>\n";
        cmd += "   <fire-all-rules/>\n";
        cmd += "</batch-execution>\n";
                
        Object object = this.context.createProducerTemplate().requestBody("direct://client", cmd);
        System.out.println( object ); 
    }

The following urls show sample script examples for jaxb, xstream and json marshalling using:

  • http://fisheye.jboss.org/browse/JBossRules/trunk/drools-camel/src/test/resources/org/drools/camel/component/jaxb.mvt?r=HEAD

  • http://fisheye.jboss.org/browse/JBossRules/trunk/drools-camel/src/test/resources/org/drools/camel/component/jaxb.mvt?r=HEAD

  • http://fisheye.jboss.org/browse/JBossRules/trunk/drools-camel/src/test/resources/org/drools/camel/component/xstream.mvt?r=HEAD

2.1. Integration with Drools Expert
2.1.1. Define a Resource
2.1.2. Event Listeners
2.2. Integration with jBPM Human Task
2.2.1. How to configure Spring with jBPM Human task

In this section we will explain the drools namespace.

Execution nodes are a context to register ksessions and kbases against for lookup.

Table 2.1. 

AttributeDescriptionRequired
idBean's id is the name to be reference from other beans.Yes

Table 2.2. 

AttributeDescriptionRequired
idBean's id is the name to be reference from other beans.Yes, but can be omitted when this tag is nested.
sourceThe path to the file containing the actual resource.Yes
typeType of the Resource as is defined in class ResourceType.No, default to DRL

When configuring a decision table resource you can nest a <drools:decisiontable-conf> tag to configure the resource.

Table 2.3. 

AttributeDescriptionRequired
input-typeDecision Table input type: XLS or CSVNo
worksheet-nameThe worksheet name to be used in the when using an XLSNo

Example 2.1. resource definition example


<drools:resource source="classpath:org/drools/IntegrationExampleTest.xls"

                 type="DTABLE">

   <drools:decisiontable-conf input-type="XLS" worksheet-name="Tables_2" />

</drools:resource>

Table 2.4. 

AttributeDescriptionRequired
idBean's id is the name to be referenced from other beans.Yes
nodeExecution-Node context to register the kbase withno

A Knowledge Base has a 0..n resources. Has two nested elements, resources and configurations. Those resources can be inner elements or references, using the "ref" attribute.

Example 2.2. kbase definition example


<drools:kbase id="kbase1" node="node1">

   <drools:resources> 

      <drools:resource source="classpath:org/drools/spring/IntegrationExampleTest.xls" 

                    type="DTABLE">

         <drools:decisiontable-conf input-type="XLS" worksheet-name="Table_2" />

      </drools:resource>



      <drools:resource ref="resource1"/>



      <drools:resource source="classpath:org/drools/container/spring/model.xsd" />

   </drools:resources>



   <drools:configuration>

       <drools:mbeans enabled="true" />



      <drools:accumulate-functions>

         <drools:accumulate-function name="func1" ref="func1Instance" />

         <drools:accumulate-function name="func1" ref="func2Instance" />

      </drools:accumulate-functions>

   </drools:configuration>



</drools:kbase>

  • advanced-process-rule-integration

    • enabled = true : false

  • multithread

    • enabled = true : false

    • max-threads = 1..n

  • mbeans

    • enabled = true : false

  • event-processing-mode

    • mode = STREAM : CLOUD

  • accumulate-functions

    • accumulate-function 0..n

      • name = String

      • ref = String

  • evaluators

    • evaluator 0..n

      • name = String

      • ref = String

  • assert-behavior

    • mode = IDENTITY : EQUALITY

Figure 2.1. Knowledge Base Configuration Options


Table 2.5. 

AttributeDescriptionRequired
idBean's id is the name to be referenced from other beans.Yes
typeis the session stateful or stateless?Yes
name No; defaults to id when omitted.
nodeExecution-Node context to register the ksession withno
listenersSpecifies the reference to the event listeners group (see 'Defining a Group of listeners' section below).no

Example 2.3. ksession definition example


<drools:ksession id="ksession1" type="stateless" 

                  name="stateless1" kbase="kbase1"/>



<drools:ksession id="ksession2" type="stateful" kbase="kbase1"/>



<drools:ksession id="ksession3" type="stateful" kbase="kbase2>



   <drools:batch>

      <drools:insert-object ref="person" />



      <drools:set-global identifier="list1">

          <bean class="java.util.ArrayList" />

      </drools:set-global>



      <drools:startProcess process-id="start fire"> 



   </drools:batch>



   <drools:configurations>

     <drools:keep-reference enabled="false" />

     <drools:clock-type type="PSEUDO"  />

   </drools:configurations>



</drools:ksession>

  • keep-reference

    • enabled = true : false

  • clock-type

    • type = REALTIME : PSEUDO

  • jpa-persistence

    • transaction-manager

      • ref = String

    • entity-manager-factory

      • ref = String

Figure 2.2. Knowledge Session Configuration Options


  • insert-object

    • ref = String (optional)

    • Anonymous bean

  • set-global

    • identifier = String (required)

    • reg = String (optiona)

    • Anonymous bean

  • fire-all-rules

    • max : n

  • fire-until-halt

  • start-process

    • parameter

      • identifier = String (required)

      • ref = String (optional)

      • Anonymous bean

  • signal-event

    • ref = String (optional)

    • event-type = String (required)

    • process-instance-id =n (optional)

Figure 2.3. Initialization Batch Commands


Example 2.4. ksession JPA configuration example


<drools:kstore id="kstore" /> <!-- provides KnowledgeStoreService implementation -->



<bean id="myEmf" class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean">

   <property name="dataSource" ref="ds" />

   <property name="persistenceUnitName" value="org.drools.persistence.jpa.local" />

</bean>



<bean id="txManager" class="org.springframework.orm.jpa.JpaTransactionManager">

   <property name="entityManagerFactory" ref="myEmf" />

</bean>

   

<drools:ksession id="jpaSingleSessionCommandService" type="stateful"  kbase="kbProcessWorkItems">

   <drools:configuration>

      <drools:jpa-persistence>

         <drools:transaction-manager ref="txManager" />

         <drools:entity-manager-factory ref="myEmf" />   

      </drools:jpa-persistence>

   </drools:configuration>

</drools:ksession>

When configuring a Event Listener resource you can nest a <drools:eventListener> tag to configure the listener.

Drools supports adding 3 types of listeners to KnowledgeSessions - AgendaListener, WorkingMemoryListener, ProcessEventListener

The drools-spring module allows you to configure these listeners to KnowledgeSessions using XML tags. These tags have identical names as the actual listener interfaces i.e., <drools:agendaEventListener....>, <drools:workingMemoryEventListener....> and <drools:processEventListener....>.

drools-spring provides features to define the listeners as standalone (individual) listeners and also to define them as a group.

Table 2.6. 

AttributeRequiredDescription
refNoA reference to another declared bean.


Example 2.5. Listener configuration example - using a bean:ref.


  <drools:resource id="resource1" type="DRL" source="classpath:org/drools/container/spring/testSpring.drl"/>



  <drools:kbase id="kbase1">

    <drools:resources>

      <drools:resource ref="resource1"/>

    </drools:resources>

  </drools:kbase>



  <bean id="mock-agenda-listener" class="org.drools.container.spring.MockAgendaEventListener" />

  <bean id="mock-wm-listener" class="org.drools.container.spring.MockWorkingMemoryEventListener" />

  <bean id="mock-process-listener" class="org.drools.container.spring.MockProcessEventListener" />



  <drools:ksession id="statefulSession" type="stateful" kbase="kbase1" node="node1">

    <drools:agendaEventListener ref="mock-agenda-listener"/>

    <drools:processEventListener ref="mock-process-listener"/>

    <drools:workingMemoryEventListener ref="mock-wm-listener"/>

  </drools:ksession>

Example 2.6. Listener configuration example - using nested bean.


  <drools:resource id="resource1" type="DRL" source="classpath:org/drools/container/spring/testSpring.drl"/>



  <drools:kbase id="kbase1">

    <drools:resources>

      <drools:resource ref="resource1"/>

    </drools:resources>

  </drools:kbase>



  <drools:ksession id="statefulWithNestedBean" type="stateful" kbase="kbase1" node="node1">



    <drools:agendaEventListener>

      <bean class="org.drools.container.spring.MockAgendaEventListener"/>

    </drools:agendaEventListener>



    <drools:workingMemoryEventListener>

      <bean class="org.drools.container.spring.MockWorkingMemoryEventListener"/>

    </drools:workingMemoryEventListener>



    <drools:processEventListener>

      <bean class="org.drools.container.spring.MockProcessEventListener"/>

    </drools:processEventListener>



  </drools:ksession>

When a listener is defined without a reference to a implementing bean and does not contain a nested bean, <drools:workingMemoryEventListener/> the underlying implementation adds the Debug version of the listener defined in the API.

The debug listeners print the corresponding Event toString message to System.err.

Example 2.7. Listener configuration example - defaulting to the debug versions provided by the Knowledge-API .


  <drools:resource id="resource1" type="DRL" source="classpath:org/drools/container/spring/testSpring.drl"/>



  <drools:kbase id="kbase1">

    <drools:resources>

      <drools:resource ref="resource1"/>

    </drools:resources>

  </drools:kbase>



  <drools:ksession id="statefulWithDefault" type="stateful" kbase="kbase1" node="node1">



    <drools:agendaEventListener/> <!-- attaches the org.drools.core.event.rule.DebugAgendaEventListener -->

    <drools:workingMemoryEventListener/> <!-- attaches the org.drools.core.event.rule.DebugWorkingMemoryEventListener -->

    <drools:processEventListener/> <!-- attaches the org.drools.core.event.DebugProcessEventListener -->



  </drools:ksession>

The drools-spring module allows you to mix and match the different declarative styles within the same KnowledgeSession. The below sample provides more clarity.

Example 2.8. Listener configuration example - mix and match of 'ref'/nested-bean/empty styles.


  <drools:resource id="resource1" type="DRL" source="classpath:org/drools/container/spring/testSpring.drl"/>



  <drools:kbase id="kbase1">

    <drools:resources>

      <drools:resource ref="resource1"/>

    </drools:resources>

  </drools:kbase>



  <bean id="mock-agenda-listener" class="org.drools.container.spring.MockAgendaEventListener" />



  <drools:ksession id="statefulWithMixAndMatchOfStyles" type="stateful" kbase="kbase1" node="node1">



     <drools:workingMemoryEventListener>

      <bean class="org.drools.container.spring.MockWorkingMemoryEventListener"/>

    </drools:workingMemoryEventListener>



    <drools:agendaEventListener ref="mock-agenda-listener"/>



    <drools:processEventListener/> <!-- attaches the org.drools.core.event.DebugProcessEventListener -->



  </drools:ksession>

It is also valid to define multiple beans of the same event listener types for a KnowledgeSession.

Example 2.9. Listener configuration example - multiple listeners of the same type.


  <drools:resource id="resource1" type="DRL" source="classpath:org/drools/container/spring/testSpring.drl"/>



  <drools:kbase id="kbase1">

    <drools:resources>

      <drools:resource ref="resource1"/>

    </drools:resources>

  </drools:kbase>



  <bean id="mock-wm-listener" class="org.drools.container.spring.MockWorkingMemoryEventListener" />



  <drools:ksession id="statefulWithMultipleSameType" type="stateful" kbase="kbase1" node="node1">



    <!-- 2 different implementations of the same listener attached to one ksession -->

    <drools:workingMemoryEventListener>

      <bean class="org.drools.container.spring.ConsoleWorkingMemoryEventListener"/>

    </drools:workingMemoryEventListener>



    <drools:workingMemoryEventListener ref="mock-wm-listener"/>



  </drools:ksession>

drools-spring allows for grouping of listeners. This is particularly useful when you define a set of listeners and want to attach them to multiple sessions. The grouping feature is also very useful, when we define a set of listeners for 'testing' and then want to switch them for 'production' use.

Table 2.7. 

AttributeRequiredDescription
IDyesUnique identifier

Note

The above mentioned child elements can be declared in any order. Only one declaration of each type is allowed in a group.

Example 2.10. Group of listeners - example


  <drools:resource id="resource1" type="DRL" source="classpath:org/drools/container/spring/testSpring.drl"/>



  <drools:kbase id="kbase1">

    <drools:resources>

      <drools:resource ref="resource1"/>

    </drools:resources>

  </drools:kbase>



  <bean id="mock-wm-listener" class="org.drools.container.spring.MockWorkingMemoryEventListener" />



  <drools:ksession id="statelessWithGroupedListeners" type="stateless" kbase="kbase1" node="node1" listeners="

mockListeners
"/>



  <drools:eventListeners id="mockListeners">

    <drools:agendaEventListener ref="mock-agenda-listener"/>

    <drools:processEventListener ref="mock-process-listener"/>

    <drools:workingMemoryEventListener ref="mock-wm-listener"/>

  </drools:eventListeners>



  <drools:eventListeners id="debugListeners">

    <drools:agendaEventListener/>

    <drools:processEventListener/>

    <drools:workingMemoryEventListener/>

  </drools:eventListeners>

This chapter describes the infrastructure used when configuring a human task server with Spring as well as a little bit about the infrastructure used when doing this.

The jBPM human task server can be configured to use Spring persistence. Example 2.11, “Configuring Human Task with Spring” is an example of this which uses local transactions and Spring's thread-safe EntityManager proxy.

The following diagram shows the dependency graph used in Example 2.11, “Configuring Human Task with Spring”.

Spring Human Task integration injection dependencies

Figure 2.4. Spring Human Task integration injection dependencies


A TaskService instance is dependent on two other bean types: a drools SystemEventListener bean as well as a TaskSessionSpringFactoryImpl bean. The TaskSessionSpringFactoryImpl bean is howerver not injected into the TaskService bean because this would cause a circular dependency. To solve this problem, when the TaskService bean is injected into the TaskSessionSpringFactoryImpl bean, the setter method used secretly injects the TaskSessionSpringFactoryImpl instance back into the TaskService bean and initializes the TaskService bean as well.

The TaskSessionSpringFactoryImpl bean is responsible for creating all the internal instances in human task that deal with transactions and persistence context management. Besides a TaskService instance, this bean also requires a transaction manager and a persistence context to be injected. Specifically, it requires an instance of a HumanTaskSpringTransactionManager bean (as a transaction manager) and an instance of a SharedEntityManagerBean bean (as a persistence context instance).

We also use some of the standard Spring beans in order to configure persistence: there's a bean to hold the EntityManagerFactory instance as well as the SharedEntityManagerBean instance. The SharedEntityManagerBean provides a shared, thread-safe proxy for the actual EntityManager.

The HumanTaskSpringTransactionManager bean serves as a wrapper around the Spring transaction manager, in this case the JpaTransactionManager. An instance of a JpaTransactionManager bean is also instantiated because of this.

Example 2.11. Configuring Human Task with Spring


<?xml version="1.0" encoding="UTF-8"?>

<beans xmlns="http://www.springframework.org/schema/beans"

       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

       xmlns:jbpm="http://drools.org/schema/drools-spring"

       xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-3.0.xsd

       http://drools.org/schema/drools-spring org/drools/container/spring/drools-spring-1.2.0.xsd">



  <!-- persistence & transactions-->

  <bean id="htEmf" class="org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean">

    <property name="persistenceUnitName" value="org.jbpm.task" />

  </bean>



  <bean id="htEm" class="org.springframework.orm.jpa.support.SharedEntityManagerBean">

    <property name="entityManagerFactory" ref="htEmf"/>

  </bean>



  <bean id="jpaTxMgr" class="org.springframework.orm.jpa.JpaTransactionManager">

    <property name="entityManagerFactory" ref="htEmf" />

    <!-- this must be true if using the SharedEntityManagerBean, and false otherwise -->

    <property name="nestedTransactionAllowed" value="true"/>

  </bean>



  <bean id="htTxMgr" class="org.drools.container.spring.beans.persistence.HumanTaskSpringTransactionManager">

    <constructor-arg ref="jpaTxMgr" />

  </bean>



  <!-- human-task beans -->



  <bean id="systemEventListener" class="org.drools.SystemEventListenerFactory" factory-method="getSystemEventListener" />



  <bean id="taskService" class="org.jbpm.task.service.TaskService" >

    <property name="systemEventListener" ref="systemEventListener" />

  </bean>



  <bean id="springTaskSessionFactory" class="org.jbpm.task.service.persistence.TaskSessionSpringFactoryImpl"

        init-method="initialize" depends-on="taskService" >

    <!-- if using the SharedEntityManagerBean, make sure to enable nested transactions -->

    <property name="entityManager" ref="htEm" />

    <property name="transactionManager" ref="htTxMgr" />

    <property name="useJTA" value="false" />

    <property name="taskService" ref="taskService" />

  </bean>



</beans>

When using the SharedEntityManagerBean instance, it's important to configure the Spring transaction manager to use nested transactions. This is because the SharedEntityManagerBean is a transactional persistence context and will close the persistence context after every operation. However, the human task server needs to be able to access (persisted) entities after operations. Nested transactions allow us to still have access to entities that otherwise would have been detached and are no longer accessible, especially when using an ORM framework that uses lazy-initialization of entities.

Also, while the TaskSessionSpringFactoryImpl bean takes an “useJTA” parameter, at the moment, JTA transactions with Spring have not yet been fully tested.

3.1. Introduction
3.2. Deployment
3.3. Configuration
3.3.1. REST/Camel Services configuration

The drools camel server (drools-camel-server) module is a war which you can deploy to execute KnowledgeBases remotely for any sort of client application. This is not limited to JVM application clients, but any technology that can use HTTP, through a REST interface. This version of the execution server supports stateless and stateful sessions in a native way.

Drools Camel Server is a war file, which can be deployed in a application server (such as JBoss AS). As the service is stateless, it is possible to have have as many of these services deployed as you need to serve the client load. Deploy on JBoss AS 4.x / Tomcat 6.x works out-of-the-box, instead some external dependencies must be added and the configuration must be changed to be deployed in JBoss AS 5

Inside the war file you will find a few XML configuration files.

The next step is configure the services that are going to be exposed through drools-server. You can modify this configuration in camel-server.xml file.


<?xml version="1.0" encoding="UTF-8"?>

<beans xmlns="http://www.springframework.org/schema/beans"

  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

  xmlns:cxf="http://camel.apache.org/schema/cxf"

  xmlns:jaxrs="http://cxf.apache.org/jaxrs"

  xsi:schemaLocation="

  http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-2.5.xsd

  http://camel.apache.org/schema/cxf http://camel.apache.org/schema/cxf/camel-cxf.xsd

  http://cxf.apache.org/jaxrs http://cxf.apache.org/schemas/jaxrs.xsd

  http://camel.apache.org/schema/spring http://camel.apache.org/schema/spring/camel-spring.xsd">



<import resource="classpath:META-INF/cxf/cxf.xml" />

<import resource="classpath:META-INF/cxf/cxf-extension-jaxrs-binding.xml"/> 

<import resource="classpath:META-INF/cxf/cxf-servlet.xml" />



  <!--

   !   If you are running on JBoss you will need to copy a camel-jboss.jar into the lib and set this ClassLoader configuration

   !  http://camel.apache.org/camel-jboss.html

   !   <bean id="jbossResolver" class="org.apache.camel.jboss.JBossPackageScanClassResolver"/>

   -->



  <!--

   !   Define the server end point.

   !   Copy and paste this element, changing id and the address, to expose services on different urls.

   !   Different Camel routes can handle different end point paths.

   -->

  <cxf:rsServer id="rsServer"  

                address="/kservice/rest"

                serviceClass="org.drools.jax.rs.CommandExecutorImpl">

    <cxf:providers>

      <bean class="org.drools.jax.rs.CommandMessageBodyReader"/>

    </cxf:providers>

  </cxf:rsServer>  



  <!-- Leave this, as it's needed to make Camel "drools" aware -->

  <bean id="droolsPolicy" class="org.drools.camel.component.DroolsPolicy" />  



  <camelContext id="camel" xmlns="http://camel.apache.org/schema/spring">    

    <!-- 

     ! Routes incoming messages from end point id="rsServer".

     ! Example route unmarshals the messages with xstream and executes against ksession1.

     ! Copy and paste this element, changing marshallers and the 'to' uri, to target different sessions, as needed.

     !-->

     

    <route>

      <from uri="cxfrs://bean://rsServer"/>

        <policy ref="droolsPolicy">

          <unmarshal ref="xstream" />       

          <to uri="drools:node1/ksession1" />

          <marshal ref="xstream" />

        </policy>

    </route>    

        

  </camelContext>

  

</beans> 

In the next xml snippet code we are creating a RESTful (JAX-RS) endpoint bound to /kservice/rest address and using org.drools.jax.rs.CommandExecutorImpl as the service implementer. This class is only used to instantiate the service endpoint because all the internal implementation is managed by Camel, and you can see in the source file that the exposed execute service must be never called.

Also a JAX-RS Provider is provided to determine if the message transported can be processed in this service endpoint.


<cxf:rsServer id="rsServer"

               address="/kservice/rest"

               serviceClass="org.drools.jax.rs.CommandExecutorImpl">

  <cxf:providers>

    <bean class="org.drools.jax.rs.CommandMessageBodyReader"/>

  </cxf:providers>

</cxf:rsServer>

Ideally this configuration doesn't need to be modified, at least the Service Class and the JAX-RS Provider, but you can add more endpoints associated to different addresses to use them in other Camel Routes.

After all this initial configuration, you can start config your own Knowledge Services.

DroolsPolicy is used to add Drools support in Camel, basically what it does is to add interceptors into the camel route to create Camel Processors on the fly and modify the internal navigation route. If you want to have SOAP support you need to create your custom Drools Policy, but it's going to be added in the next release.

But you don’t need to know more internal details, only instantiate this bean:


<bean id="droolsPolicy" class="org.drools.camel.component.DroolsPolicy" />

The next is create the camel route that will have the responsibility to execute the commands sent through JAX-RS. Basically we create a route definition associated with the JAX-RS definition as the data input, the camel policy to be used and inside the “execution route” or ProcessorDefinitions. As you can see, we set XStream as the marshaller/unmarshaller and the drools execution route definition


<camelContext id="camel" xmlns="http://camel.apache.org/schema/spring">

  <route>

    <from uri="cxfrs://bean://rsServer"/>

    <policy ref="droolsPolicy">

      <unmarshal ref="xstream" />

      <to uri="drools:node1/ksession1" />

      <marshal ref="xstream" />

    </policy>

  </route>

</camelContext>

The drools endpoint creation has the next arguments

<!-- XML : generated by JHighlight v1.0 (http://jhighlight.dev.java.net) -->
<span class="xml_tag_symbols">&lt;</span><span class="xml_tag_name">to</span><span class="xml_plain">&nbsp;</span><span class="xml_attribute_name">uri</span><span class="xml_tag_symbols">=</span><span class="xml_attribute_value">&quot;drools:{1}/{2}&quot;</span><span class="xml_plain">&nbsp;</span><span class="xml_tag_symbols">/&gt;</span><span class="xml_plain"></span><br />

  1. Execution Node identifier that is registered in the CamelContext

  2. Knowledge Session identifier that was registered in the Execution Node with identifier {1}

Both parameters are configured in knowledge-services.xml file.

The next step is create the Knowledge Sessions that you are going to use.


<beans xmlns="http://www.springframework.org/schema/beans"

       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

       xmlns:drools="http://drools.org/schema/drools-spring"

       xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-2.0.xsd

                           http://drools.org/schema/drools-spring http://drools.org/schema/drools-spring.xsd">



   <drools:execution-node id="node1" />



   <drools:kbase id="kbase1" node="node1">

      <drools:resources>

         <drools:resource  type="XSD" source="classpath:model.xsd"/>

         <drools:resource  type="DRL" source="classpath:test.drl"/>

      </drools:resources>

   </drools:kbase>



   <drools:ksession id="ksession1" type="stateless" kbase="kbase1" node="node1"/>

</beans>

The execution-node is a context or registered kbases and ksessions, here kbase1 and ksession1 are planed in the node1 context. The kbase itself consists of 2 knowledge definitions, a drl and an xsd. The Spring documentation contains a lot more information on configuring these knowledge services.

With drools-server war unzipped you should be able to see a test.jsp and run it. This example just executes a simple "echo" type application. It sends a message to the rule server that pre-appends the word "echo" to the front and sends it back. By default the message is "Hello World", different messages can be passed using the url parameter msg - test.jsp?msg="My Custom Message".

Under the hood the jsp invokes the Test.java class, this then calls out to Camel which is where the meet happens. The camel-client.xml defines the client with just a few lines of xml:


<camelContext id="camel" xmlns="http://camel.apache.org/schema/spring">

   <route>

      <from uri="direct://kservice"/>

      <policy ref="droolsPolicy">

         <to uri="cxfrs://http://localhost:8080/drools-server-app/kservice/rest"/>

      </policy>

   </route>         

 </camelContext>

"direct://kservice" is just a named hook, allowing java to grab a reference and push data into it. In this example the data is already in xml, so we don't need to add any DataFormat's to do the marshalling. The DroolsPolicy adds some smarts to the route and you'll see it used on the server side too. If JAXB or XStream were used, it would inject custom paths and converters, it can also set the ClassLoader too on the server side, on the client side it automatically unwraps the Response object.

The rule itself can be found here: test.drl. Notice the type Message is declared part of the drl and is thus not present on the Classpath.

declare Message
   text : String
end
    
  
rule "echo" dialect "mvel"
when
   $m : Message();
then
   $m.text = "echo:" + $m.text;
end
4.1. API
4.1.1. XStream
4.1.2. JSON
4.1.3. JAXB
4.2. Commands supported
4.2.1. BatchExecutionCommand
4.2.2. InsertObjectCommand
4.2.3. RetractCommand
4.2.4. ModifyCommand
4.2.5. GetObjectCommand
4.2.6. InsertElementsCommand
4.2.7. FireAllRulesCommand
4.2.8. StartProcessCommand
4.2.9. SignalEventCommand
4.2.10. CompleteWorkItemCommand
4.2.11. AbortWorkItemCommand
4.2.12. QueryCommand
4.2.13. SetGlobalCommand
4.2.14. GetGlobalCommand
4.2.15. GetObjectsCommand

XML marshalling/unmarshalling of the Drools Commands requires the use of special classes, which are going to be described in the following sections.

The following urls show sample script examples for jaxb, xstream and json marshalling using:

To use the XStream commands marshaller you need to use the DroolsHelperProvider to obtain an XStream instance. We need to use this because it has the commands converters registered.

JSON API to marshalling/unmarshalling is similar to XStream API:

There are two options for using JAXB, you can define your model in an XSD file or you can have a POJO model. In both cases you have to declare your model inside JAXBContext, and in order to do that you need to use Drools Helper classes. Once you have the JAXBContext you need to create the Unmarshaller/Marshaller as needed.

With your model defined in a XSD file you need to have a KnowledgeBase that has your XSD model added as a resource.

To do this, the XSD file must be added as a XSD ResourceType into the KnowledgeBuilder. Finally you can create the JAXBContext using the KnowledgeBase created with the KnowledgeBuilder



Options xjcOpts = new Options();

xjcOpts.setSchemaLanguage(Language.XMLSCHEMA);

JaxbConfiguration jaxbConfiguration = KnowledgeBuilderFactory.newJaxbConfiguration( xjcOpts, "xsd" );

kbuilder.add(ResourceFactory.newClassPathResource("person.xsd", getClass()), ResourceType.XSD, jaxbConfiguration);

KnowledgeBase kbase = kbuilder.newKnowledgeBase();


List<String> classesName = new ArrayList<String>();

classesName.add("org.drools.compiler.test.Person");

   

JAXBContext jaxbContext = KnowledgeBuilderHelper.newJAXBContext(classesName.toArray(new String[classesName.size()]), kbase);

In this case you need to use DroolsJaxbHelperProviderImpl to create the JAXBContext. This class has two parameters:

  1. classNames: A List with the canonical name of the classes that you want to use in the marshalling/unmarshalling process.

  2. properties: JAXB custom properties

List<String> classNames = new ArrayList<String>();

classNames.add("org.drools.compiler.test.Person");

JAXBContext jaxbContext = DroolsJaxbHelperProviderImpl.createDroolsJaxbContext(classNames, null);

Marshaller marshaller = jaxbContext.createMarshaller();

Currently, the following commands are supported:

Note

In the next snippets code we are going to use a POJO org.drools.compiler.test.Person that has two fields

Note

In the next examples, to marshall the commands we have used the next snippet codes:

OSGi is a dynamic module system for declarative services. So what does that mean? Each jar in OSGi is called a bundle and has it's own ClassLoader. Each bundle specifies the packages it exports (makes publicly available) and which packages it imports (external dependencies). OSGi will use this information to wire the ClassLoaders of different bundles together; the key distinction is you don't specify what bundle you depend on, or have a single monolithic classpath, instead you specify your package import and version and OSGi attempts to satisfy this from available bundles.

It also supports side by side versioning, so you can have multiple versions of a bundle installed and it'll wire up the correct one. Further to this Bundles can register services for other bundles to use. These services need initialisation, which can cause ordering problems - how do you make sure you don't consume a service before its registered? OSGi has a number of features to help with service composition and ordering. The two main ones are the programmatic ServiceTracker and the xml based Declarative Services. There are also other projects that help with this; Spring DM, iPOJO, Gravity.

Each of the Drools factories is now also available as a FactoryService interface. You can either have OSGi inject those into a pojo, or retrieve them yourself from OSGi.

The following modules should work with OSGi;

The following Services can be located as OSGi Bundles

I'll cover injection here. The below example injects the KnowledgeBuilderFactoryService, KnowledgeBaseFactoryService and ResourceFactoryService into the TestComponent pojo.

Example 5.1. OSGi Declarative Services


<scr:component xmlns:scr="http://www.osgi.org/xmlns/scr/v1.1.0">

<implementation class="testosgi.TestComponent"/>



<reference bind="setKnowledgeBaseFactoryService"

           unbind="unsetKnowledgeBaseFactoryService"

           interface="org.drools.KnowledgeBaseFactoryService"

           />



<reference bind="setResourceFactoryService"

           unbind="unsetResourceFactoryService"

           interface="org.drools.core.io.ResourceFactoryService"

           />



<reference bind="setKnowledgeBuilderFactoryService"

           unbind="unsetKnowledgeBuilderFactoryService"

           interface="org.drools.core.builder.KnowledgeBuilderFactoryService"

           target="(org.drools.compiler.compiler.DecisionTableProvider=true)"  />

</scr:component>

The TestComponent will only be activated when all of the referenced services are available and injected into the pojo. You'll also notice the "target" attribute for the KnowledgeBuilderFactoryService. The reason for this is that OSGi DS has no built in way to declaratively say which optional services must be present to satisfy your component. As a work around I made any Drools service that has optional services set a property if/when the optional service is available. Filters can then be applied, via the target attribute, to make sure the Service is in a desired state before consuming it. And that is pretty much it :)

Example 5.2. Basic Rule Compilation

ServiceReference serviceRef = bundleContext.getServiceReference( ServiceRegistry.class.getName() );
ServiceRegistry registry = (ServiceRegistry) bundleContext.getService( serviceRef );

KnowledgeBuilderFactoryService knowledgeBuilderFactoryService = registry.get( KnowledgeBuilderFactoryService.class );
        
KnowledgeBaseFactoryService knowledgeBaseFactoryService = registry.get( KnowledgeBaseFactoryService.class );
ResourceFactoryService resourceFactoryService = registry.get( ResourceFactoryService.class );

KnowledgeBuilderConfiguration kbConf = knowledgeBuilderFactoryService.newKnowledgeBuilderConfiguration( null,
                                                                                                        getClass().getClassLoader() );

KnowledgeBuilder kbuilder = knowledgeBuilderFactoryService.newKnowledgeBuilder( kbConf );
ResourceFactoryService resource = resourceFactoryService;
kbuilder.add( resource.newByteArrayResource( string.getBytes() ),
              ResourceType.DRL );

if ( kbuilder.hasErrors() ) {
     System.out.println( kbuilder.getErrors() );
     throw new RuntimeException( kbuilder.getErrors().toString() );
}

KnowledgeBaseConfiguration kbaseConf = knowledgeBaseFactoryService.newKnowledgeBaseConfiguration( null,
                                                                                                  getClass().getClassLoader() );

KnowledgeBase kbase = knowledgeBaseFactoryService.newKnowledgeBase( kbaseConf );
kbase.addKnowledgePackages( kbuilder.getKnowledgePackages() );
StatefulKnowledgeSession ksession = kbase.newStatefulKnowledgeSession();

Example 5.3. Decision Table Example

ServiceReference serviceRef = bundleContext.getServiceReference( ServiceRegistry.class.getName() );
ServiceRegistry registry = (ServiceRegistry) bundleContext.getService( serviceRef );

KnowledgeBuilderFactoryService knowledgeBuilderFactoryService = registry.get( KnowledgeBuilderFactoryService.class );
KnowledgeBaseFactoryService knowledgeBaseFactoryService = registry.get( KnowledgeBaseFactoryService.class );
ResourceFactoryService resourceFactoryService = registry.get( ResourceFactoryService.class );

KnowledgeBaseConfiguration kbaseConf = knowledgeBaseFactoryService.newKnowledgeBaseConfiguration( null,
                                                                                                  getClass().getClassLoader() );

KnowledgeBuilderConfiguration kbConf = knowledgeBuilderFactoryService.newKnowledgeBuilderConfiguration( null,
                                                                                                        getClass().getClassLoader() );
KnowledgeBuilder kbuilder = knowledgeBuilderFactoryService.newKnowledgeBuilder( kbConf );
kbuilder.add( resourceFactoryService.newClassPathResource( "changeset1Test.xml",
                                                           Dummy.class ),
              ResourceType.CHANGE_SET );

kbaseConf = knowledgeBaseFactoryService.newKnowledgeBaseConfiguration( null,
                                                                       getClass().getClassLoader() );
KnowledgeBase kbase = knowledgeBaseFactoryService.newKnowledgeBase( kbaseConf );
kbase.addKnowledgePackages( kbuilder.getKnowledgePackages() );

StatefulKnowledgeSession ksession = kbase.newStatefulKnowledgeSession();

6.1. Introduction
6.1.1. Enabling JMX monitoring in a Drools application
6.1.2. Installing and running the RHQ/JON plugin

The Drools engine supports runtime monitoring through JMX standard MBeans. These MBeans expose configuration and metrics data, from live knowledge bases and sessions, to internal details like rule execution times. Any JMX compatible console can be used to access that data. This chapter details how to use RHQ/JON to do it, but similar steps can be used for any other console.

To enable JMX monitoring in a Drools application, it is necessary to enable remote monitoring in the JVM. There are several tutorials on how to do that in the internet, but we recommend that you check the documentation of your specific JVM. Using the Oracle/Sun JVM, it can be as simple as running the engine with a few command line system properties.

For instance, to enable remote monitoring on port 19988 with disabled authentication (should be only used for tests/demos, as in production authentication should be enabled), just run the application with the following command line parameters:


 -Dcom.sun.management.jmxremote.port=19988 -Dcom.sun.management.jmxremote.ssl=false -Dcom.sun.management.jmxremote.authenticate=false

The second step is to enable the Drools MBeans. As any Drools configuration, that can be done by setting a system property, or adding the property to a configuration file, or using the API.

To enable it in the command line, use:


 -Ddrools.mbeans=enabled

To enable id using the API, use:


 KnowledgeBaseConfiguration conf = ... 

 conf.setOption( MBeansOption.ENABLED );

The following sequence of steps can be used to configure JON to monitor a Drools application:

  1. Download the JON server and agent.

  2. Download Drools plugin included in the "Drools and jBPM tools" bundle (http://www.jboss.org/drools/downloads.html).

  3. Install server, agent, and the plugin.

  4. Check that the server is running, agent is running and plugin is installed.

  5. Execute the drools application [see details in the previous section].

  6. On the agent console, type "discovery" command for the agent to find the drools application, which it will find on port 19988.

  7. On JON console, click on auto-discovery queue.

  8. Select the JMX Server process that is showing there, running on port 19988.

  9. Click import.

  10. Click on Resources->servers.

  11. Click on the JMX Server.

  12. Under JMXServer on the left hand side, you have Drools Service.