Accelerate your business

2013/06/11

Clustering in jBPM v6

Clustering in jBPM v5 was not an easy task, there where several known issues that had to be resolved on client (project that was implementing solution with jBPM) side, to name few:

session management - when to load/dispose knowledge session
timer management - required to keep knowledge session active to fire timers

This is not the case any more in version 6 where several improvements made their place in code base, for example new module that is responsible for complete session management was introduced - jbpm runtime manager. More on runtime manager in next post, this one focuses on how clustered solution might look like. First of all lets start with all the important pieces jbpm environment consists of:

asset repository - VFS based repository backed with GIT - this is where all the assets are stored during authoring phase
jbpm server that includes JBoss AS7 with deployed jbpm console (bpm focused web application) of kie-wb (fully features web application that combines BPM and BRM worlds)
data base - backend where all the state date is kept (process instances, ksessions, history log, etc)

Repository clustering

Asset repository is GIT backed virtual file system (VFS) that keeps all the assets (process definitions, rules, data model, forms, etc) is reliable and efficient way. Anyone who used to work with GIT understands perfectly how good it is for source management and what else assets are if not source code?

So if that is file system it resides on the same machine as the server that uses it, that enforces it to be kept in sync between all servers of a cluster. For that jbpm makes use of two well know open source projects:

Zookeeper is responsible for gluing all parts together where Helix is cluster management component that registers all cluster details (cluster itself, nodes, resources).

So this two components are utilized by the runtime environment on which jbpm v6 is based on:

kie-commons - provides VFS implementation and clustering
uber fire framework - provides backbone of the web applications

So let's take a look at what we need to do to setup cluster of our VFS:

Get the software

download Apache Zookeeper (note that 3.3.4 and 3.3.5 are the only versions that were currently tested so make sure you get the correct version)
download Apache Helix (note that version that was tested was 0.6.1)

Install and configure

unzip Apache Zookeeper into desired location - ( from now one we refer to it as zookeeper_home)
go to zookeeper_home/conf and make a copy of zoo_sample.conf to zoo.conf
edit zoo.conf and adjust settings if needed, these two are important in most of the cases:

# the directory where the snapshot is stored.
dataDir=/tmp/zookeeper
# the port at which the clients will connect
clientPort=2181

unzip Apache helix to into desired location (from now one we refer to it as helix_home)

Setup cluster

Now we have all the software available locally so next step is to configure the cluster itself. We start with start of the Zookeeper server that will be master of the configuration of the cluster:

go to zookeeper_home/bin
execute following command to start zookeeper server:

sudo ./zkServer.sh start

zookeeper server should be started, if the server fails to start make sure that the data directory defined in zoo.conf file exists and is accessible
all zookeeper activities can be viewed zookeeper_home/bin/zookeeper.out

To do so, Apache Helix provides utility scripts that can be found in helix_home/bin.

go to helix_home/bin
create cluster

./helix-admin.sh --zkSvr localhost:2181 --addCluster jbpm-cluster

add nodes to the cluster

node 1

./helix-admin.sh --zkSvr localhost:2181 --addNode jbpm-cluster nodeOne:12345

node2

./helix-admin.sh --zkSvr localhost:2181 --addNode jbpm-cluster nodeTwo:12346

add as many nodes as you will have cluster members of jBPM server (in most cases number of application servers in the cluster)

NOTE: nodeOne:12345 is the unique identifier of the node, that will be referenced later on when configuring application severs, although it looks like host and port number it is use to identify uniquely logical node.

add resources to the cluster

./helix-admin.sh --zkSvr localhost:2181

--addResource jbpm-cluster vfs-repo 1 LeaderStandby AUTO_REBALANCE

rebalance cluster to initialize it

./helix-admin.sh --zkSvr localhost:2181 --rebalance jbpm-cluster vfs-repo 2

start the Helix controller to manage the cluster

./run-helix-controller.sh --zkSvr localhost:2181

--cluster jbpm-cluster 2>&1 > /tmp/controller.log &

Values given above are just examples and can be changed according to the needs:

cluster name: jbpm-cluster

node name: nodeOne:12345, nodeTwo:12346

resource name: vfs-repo

zkSvr value must match Zookeeper server that is used.

Prepare data base

Before we start with application server configuration data base needs to be prepared, for this example we use PostgreSQL data base. jBPM server will create all required tables itself by default so there is no big work required for this but some simple tasks must be done before starting the server configuration.

Create data base user and data base

First of all PostgreSQL data base needs to be installed, next user needs to be created on the data base that will own the jbpm schema, in this example we use:

user name: jbpm

password: jbpm

Once the user is ready, data base can be created, and again for the example jbpm is chosen for the data base name.

NOTE: this information (username, password, data base name) will be used later on in application server configuration.

Create Quartz tables

Lastly Quartz related tables must be created, to do so best is to utilize the data base scripts provided with Quartz distribution, jbpm uses Quartz 1.8.5. DB scripts are usually located under QUARTZ_HOME/docs/dbTables.

Create quartz definition file

Quartz configuration that will be used by the jbpm server needs to accomodate the needs of the environment, as this guide is about to show the basic setup obviously it will not cover all the needs but will allow for further improvements.

Here is a sample configuration used in this setup:

#============================================================================

# Configure Main Scheduler Properties

#============================================================================

org.quartz.scheduler.instanceName = jBPMClusteredScheduler

org.quartz.scheduler.instanceId = AUTO

#============================================================================

# Configure ThreadPool

#============================================================================

org.quartz.threadPool.class = org.quartz.simpl.SimpleThreadPool

org.quartz.threadPool.threadCount = 5

org.quartz.threadPool.threadPriority = 5

#============================================================================

# Configure JobStore

#============================================================================

org.quartz.jobStore.misfireThreshold = 60000

org.quartz.jobStore.class=org.quartz.impl.jdbcjobstore.JobStoreCMT

org.quartz.jobStore.driverDelegateClass=org.quartz.impl.jdbcjobstore.PostgreSQLDelegate

org.quartz.jobStore.useProperties=false

org.quartz.jobStore.dataSource=managedDS

org.quartz.jobStore.nonManagedTXDataSource=notManagedDS

org.quartz.jobStore.tablePrefix=QRTZ_

org.quartz.jobStore.isClustered=true

org.quartz.jobStore.clusterCheckinInterval = 20000

#============================================================================

# Configure Datasources

#============================================================================

org.quartz.dataSource.managedDS.jndiURL=jboss/datasources/psjbpmDS

org.quartz.dataSource.notManagedDS.jndiURL=jboss/datasources/quartzNotManagedDS

Configure JBoss AS 7 domain

1. Create JDBC driver module - for this example PostgreSQL

a) go to JBOSS_HOME/modules directory (on EAP JBOSS_HOME/modules/system/layers/base)

b) create module folder org/postgresql/main

c) copy postgresql driver jar into the module folder (org/postgresql/main) as postgresql-jdbc.jar name

d) create module.xml file inside module folder (org/postgresql/main) with following content:

<resource-root path="postgresql-jdbc.jar"/>

</resources>

</dependencies>

</module>

2. Configure data sources for jbpm server

a) go to JBOSS_HOME/domain/configuration

b) edit domain.xml file

for simplicity sake we use default domain configuration which uses profile "full" that defines two

server nodes as part of main-server-group

c) locate the profile "full" inside the domain.xml file and add new data sources

main data source used by jbpm

<datasource jndi-name="java:jboss/datasources/psjbpmDS"

pool-name="postgresDS" enabled="true" use-java-context="true">

<connection-url>jdbc:postgresql://localhost:5432/jbpm</connection-url>

<driver>postgres</driver>

<user-name>jbpm</user-name>

</security>

</datasource>

additional data source for quartz (non managed pool)

<datasource jta="false" jndi-name="java:jboss/datasources/quartzNotManagedDS"

pool-name="quartzNotManagedDS" enabled="true" use-java-context="true">

<connection-url>jdbc:postgresql://localhost:5432/jbpm</connection-url>

<driver>postgres</driver>

<user-name>jbpm</user-name>

</security>

</datasource>

defined the driver used for the data sources

<xa-datasource-class>org.postgresql.xa.PGXADataSource</xa-datasource-class>

</driver>

3. Configure security domain

a) go to JBOSS_HOME/domain/configuration

b) edit domain.xml file

for simplicity sake we use default domain configuration which uses profile "full" that defines two

server nodes as part of main-server-group

c) locate the profile "full" inside the domain.xml file and add new security domain to define security

domain for jbpm-console (or kie-wb) - this is just a copy of the "other" security domain defined

there by default

<security-domain name="jbpm-console-ng" cache-type="default"> <authentication>

<login-module code="Remoting" flag="optional">

<module-option name="password-stacking" value="useFirstPass"/>

</login-module>

<login-module code="RealmDirect" flag="required">

<module-option name="password-stacking" value="useFirstPass"/>

</login-module>

</authentication>

</security-domain>

for kie-wb application, simply replace jbpm-console-ng with kie-ide as name of the security domain.

4. Configure server nodes

a) go to JBOSS_HOME/domain/configuration

b) edit host.xml file

c) locate servers that belongs to "main-server-group" in host.xml file and add following system

properties:

property name	property value	comments
org.uberfire.nio.git.dir	/home/jbpm/node[N]/repo	location where the VFS asset repository will be stored for the node[N]
org.quartz.properties	/jbpm/quartz-definition.properties	absolute file path to the quartz definition properties
jboss.node.name	nodeOne	unique node name within cluster (nodeOne, nodeTwo, etc)
org.uberfire.cluster.id	jbpm-cluster	name of the helix cluster
org.uberfire.cluster.zk	localhost:2181	location of the zookeeper server
org.uberfire.cluster.local.id	nodeOne_12345	unique id of the helix cluster node, note that ':' is replaced with '_'
org.uberfire.cluster.vfs.lock	vfs-repo	name of the resource defined on helix cluster
org.uberfire.nio.git.daemon.port	9418	port used by the GIT repo to accept client connections, must be unique for each cluster member
org.uberfire.nio.git.ssh.port	8001	port used by the GIT repo to accept client connections (over ssh), must be unique for each cluster member
org.uberfire.nio.git.daemon.host	localhost	host used by the GIT repo to accept client connections, in case cluster members run on different machines this property must be set to actual host name instead of localhost otherwise synchronization won't work
org.uberfire.nio.git.ssh.host	localhost	host used by the GIT repo to accept client connections (over ssh), in case cluster members run on different machines this property must be set to actual host name instead of localhost otherwise synchronization won't work
org.uberfire.metadata.index.dir	/home/jbpm/node[N]/index	location where index for search will be created (maintained by Apache Lucene)
org.uberfire.cluster.autostart	false	delays VFS clustering until the application is fully initialized to avoid conficts when all cluster members create local clones

examples for the two nodes:

nodeOne

<system-properties>

<property name="org.uberfire.nio.git.dir" value="/tmp/jbpm/nodeone"

boot-time="false"/>

<property name="org.quartz.properties"

value="/tmp/jbpm/quartz/quartz-db-postgres.properties" boot-time="false"/>

<property name="org.uberfire.cluster.id" value="jbpm-cluster"

boot-time="false"/>

<property name="org.uberfire.cluster.zk" value="localhost:2181"

boot-time="false"/>

<property name="org.uberfire.cluster.local.id" value="nodeOne_12345"

boot-time="false"/>

<property name="org.uberfire.cluster.vfs.lock" value="vfs-repo"

boot-time="false"/>

</system-properties>

nodeTwo

<system-properties>

<property name="org.uberfire.nio.git.dir" value="/tmp/jbpm/nodetwo"

boot-time="false"/>

<property name="org.quartz.properties"

value="/tmp/jbpm/quartz/quartz-db-postgres.properties" boot-time="false"/>

<property name="org.uberfire.cluster.id" value="jbpm-cluster"

boot-time="false"/>

<property name="org.uberfire.cluster.zk" value="localhost:2181"

boot-time="false"/>

<property name="org.uberfire.cluster.local.id" value="nodeTwo_12346"

boot-time="false"/>

<property name="org.uberfire.cluster.vfs.lock" value="vfs-repo"

boot-time="false"/>

</system-properties>

NOTE: since this example runs on single node host properties for ssh and git daemons are omitted.

Since repository synchronization is done between git servers make sure that GIT daemons are active (and properly configured - host name and port) on every cluster member.

5. Create user(s) and assign it to proper roles on application server

Add application users

In previous step security domain has been created so jbpm console (or kie-wb) users could be authenticated while logging on. Now it's time to add some users to be able to logon to the application once it's deployed. To do so:

a) go to JBOSS_HOME/bin

b) execute ./add-user.sh script and follow the instructions on the screen

- use Application realm not management

- when asked for roles, make sure you assign at least:

for jbpm-console: jbpm-console-user

for kie-wb: kie-user

add as many users you need, same goes for roles, listed above are required to be authorized to use the web application.

Add management (of application server) user

To be able to manage the application server as domain, we need to add administrator user, it's similar to what was defined for adding application users but the realm needs to be management

a) go to JBOSS_HOME/bin

b) execute ./add-user.sh script and follow the instructions on the screen

- use Management realm not application

Application server should be now ready to be used, so let's start the domain:

JBOSS_HOME/bin/domain.sh

after few seconds (it's still empty servers) you should be able to access both server nodes on following locations:

nodeOne: http://localhost:8080

nodeTwo: http://localhost:8230

administration console: http://localhost:9990/console

the port offset is configurable in host.xml for given server.

Deploy application - jBPM console (or kie-wb)

Now it's time to prepare and deploy application, either jbpm-console or kie-wb. As by default both application comes with predefined persistence that uses ExampleDS from AS7 and H2 data base there is a need to alter this configuration to use PostgreSQL data base instead.

Required changes in persistence.xml

change jta-data-source name to match one defined on application server

java:jboss/datasources/psjbpmDS

change hibernate dialect to be postgresql

org.hibernate.dialect.PostgreSQLDialect

Application build from source

If the application is built from source then you need to edit persistence.xml file that is located under:

jbpm-console-ng/jbpm-console-ng-distribution-wars/src/main/jbossas7/WEB-INF/classes/META-INF/

next rebuild the jbpm-distribution-wars module to prepare deployable package - once that is named:

jbpm-console-ng-jboss-as7.0.war

Deployable package downloaded

In case you have deployable package downloaded (which is already a war file) you need to extract it change the persistence.xml located under:

WEB-INF/classes/META-INF

once the file is edited and contains correct values to work properly with PostgreSQL data base application needs to be repackaged:

NOTE: before repackaging make use that previous war is not in the same directory otherwise it will be packaged into new war too.

jar -cfm jbpm-console-ng.war META-INF/MANIFEST.MF *

IMPORTANT: make sure that you include the same manifest file that was in original war file as it contains valuable entires.

To deploy application logon as management user into administration console of the domain and add new deployments using Runtime view of console. Once the deployment is added to the domain, assign it to the right server group - in this example we used main-server-group it will be default enable this deployment on all servers within that group - meaning deploy it on the servers. This will take a while and after successful deployment you should be able to access jbpm-console (or kie-wb) on following locations:

console on server one: http://localhost:8080/jbpm-console-ng

console on server one: http://localhost:8230/jbpm-console-ng

the context root (jbpm-console-ng) depends on the name of the war file that was deployed so if the filename will be jbpm-console-ng-jboss7.war then the context root will be jbpm-console-ng-jboss7. Same rule apply to kie-wb deployment.

And that's it - you should have fully operational jbpm cluster environment!!!

Obviously in normal scenarios you would like to hide the complexity of different urls to the application from end users (like putting in front of them a load balancer) but I explicitly left that out of this example to show proper behavior of independent cluster nodes.

Next post will go into details on how different components play smoothly in cluster, to name few:

failover - in case cluster node goes down
timer management - how does timer fire in cluster environment
session management - auto reactivation of session on demand
etc

As we are still in development mode, please share your thoughts on what would you like to see in cluster support for jBPM, your input is most appreciated!

Update:
There was a change in naming of system properties since the article was written so for those that configured it already for 6.0.0.Final there will be a need to adjust name of following system properties:

org.kie.nio.git.dir -> org.uberfire.nio.git.dir
org.kie.nio.git.daemon.port -> org.uberfire.nio.git.daemon.port
org.kie.kieora.index.dir -> org.uberfire.metadata.index.dir
org.uberfire.cluster.autostart - new parameter

Table above already contains proper values for 6.0.0.Final

2013/01/04

jBPM web designer runs on VFS

As part of efforts for jBPM and Drools version 6.0 web designer is going through quite few enhancements too. One of the major features is to provide flexible mechanism to persist modeled processes (and other assets that relate to them such as forms, process image, etc) a.k.a assets even without being embedded in Drools Guvnor.
So let's start with the main part here - what does it mean flexible mechanism to persists assets? To answers this let's look at what is currently (jBPM 5.x) available:

designer by default runs in embedded mode inside Drools Guvnor
designer stores all assets inside Drools Guvnor JCR repository
designer can run in standalone mode but only as modeling tool without capabilities to store assets

So as listed above there is only one option to persist assets - inside Drools Guvnor. In most of the cases this is good enough or even desired but there are quite few situation where modeling capabilities are required to be delivered with the custom application and including complete Drools Guvnor could be too much.

That leads us to the flexible mechanism implemented - designer was equipped with Repository interface that is considered entry point to interact with underlying storage. Designer by default comes with Virtual File System based repository that provides:

default implementation that supports

simple (local) file system repository
git based repository

allows for pluggable VFS provider implementations
is based on standards - java NIO2

Extensions to what is delivered out of the box can be done in one of the following ways:

if VFS based repository is not what user needs an alternative implementation of the Repository interface can be provided, e.g data base
if VFS is what user is looking for but neither local file system nor git is the right implementation additional providers can be developed

Let's look little bit deeper into what these new features are and how users will benefit from them.

1. It's based on Java 7 NIO2

VFS support provides is based on Java SE 7 NIO2 but does not require Java 7 to run as it comes with backport implementation of selected parts of NIO2 that are required

2. Different providers for Virtual File System

The simplest option is to use designer with local file system storage that will simply utilize file system on which designer is running. As most likely it will provide the best performance it leaves user with rather limited options when it comes to clustering, distributed environments or backups.

Next option that personally would recommend is to utilize GIT as underlying storage. People that works with GIT in their software development projects will most likely notice quite few advantages as in the end process definitions are more like source code that can be versioned, developed in parallel (branching) and included in some sort of release cycle.

3. Save process directly in designer editor

Designer now allows users to save process directly from the editor which will store the svg process content as well with just one click!

4. Repository menu

With the repository designer provides a simple UI menu to navigate through the repository and perform basic operations such as:

open processes in designer editor
create assets and directories
copy/move assets
delete assets and directories
preview files

This menu is intended to be seen as basic file system browser utilized more in a standalone mode as when integrated with jBPM console-ng, guvnor-ng (UberFire) more advanced options will be delivered in this area.

5. Simpler integration with jBPM console and Drools Guvnor

Both jBPM console and Drools Guvnor are going to be "refreshed" for version 6.0 and thus integration between these components and designer will be simplified as they all will be unified on the repository level, meaning single repository can be shared across all these three components.

That will be all for a brief introduction but certainly not all in this topic. Expect more to come as soon as preview will be released on how to configure different repositories and more updates on git based repository on how to make best of it.

Your comments are more than welcome as they can help to make designer best modeling tool out there :)

Known limitation

Currently git based repository does not support move of assets and directories as atomic operation which means that preferred is to copy first and then delete.

2012/10/19

Dispose session in CMT environment

Since jBPM5 is flexible process engine it can be deployed in various flavours. One of them is to embed it into you enterprise application running on application server, regardless of its vendor (JBoss, WebSphere, WebLogic, etc).
One option among many others is to make use of it as part of you business logic implemented as EJB. If you choose to use bean managed transaction (BMT) you do not need to take any additional steps as your business logic maintains transaction boundaries. Although when you use container managed transaction (CMT) situation is little bit different as it is container (application server) responsibility to manage transaction.

Before we jump into details what needs to be done for CMT based application, let's mention one important and common for both types (BMT and CMT) practice:

Session must be disposed outside transaction, meaning transaction must be committed/rolledback before session could be disposed.

Obviously this applies to situation when session should be disposed as part of business logic, for instance with session per process instance architecture this could be desired. But not when we have single centralized session.

If session will be disposed before transaction is completed, exception will be thrown on transaction completion, complaining that session is already disposed:

IllegalStateException
("Illegal method call. This session was previously disposed.")

Having this in mind, let's take a look at how this can be done in CMT based implementations. Since we do not control transaction how we could dispose session after transaction is completed?
A simple answer to this is to use dedicated Command that will register transaction synchronization object that will be called on transaction completion so we could safely dispose session.

Here is an example of such command's execute method:

public Void execute(Context context) {

final StatefulKnowledgeSession ksession =
                ((KnowledgeCommandContext)context).getStatefulKnowledgesession();
try {
      TransactionManager tm =
      (TransactionManager) new InitialContext().lookup(tmLookupName);
      tm.getTransaction().registerSynchronization(new Synchronization() {

          @Override
          public void beforeCompletion() {
              // not used here
          }

          @Override
          public void afterCompletion(int arg0) {
        ksession.dispose();
          }
      });
} catch (Exception e) {
      e.printStackTrace();
}

    return null
}

so, instead of calling default ksession.dispose() method at the end of you business logic, simply call ksession.execute(new CMTDisposeCommand());
That will ensure session is disposed as part of transaction completion.

Here is complete CMT dispose command

2012/10/05

jBPM5 Developer Guide book is on its way!!!

For those that keep an eye on jBPM user forum it won't be any surprise as it was already announced and sometime ago even request for comments was posted. So, I would like to emphasise once again - a brand new book for developers about jBPM5 and Drools by Mauricio Salatino a.k.a. Salaboy is coming.

I had a pleasure to slightly help with this book by reviewing it. I admit this is an excellent reading for those that are new to jBPM5 (and Drools) world as well as to those that are already familiar with it and are planning to put it into mission critical systems. Mauricio provides not only guidelines about the engine but goes beyond that by elaborating about different approaches on how to best utilize these frameworks.

It is much more than title would suggest, it is not only about jBPM5 (but main focus is on the business processes) but it covers topics like:

BPMN2 specification
Human tasks and WS-HT specification
Rule engine (Drools Expert)
Complex Event Processing (Drools Fusion)
Tooling (jBPM Web Designer, eclipse modeller, etc)
Centrailzed repository (Drools Guvnor)

Book is equipped with number of real life examples including source code that allows reader to follow all the content put in every chapter.

In conclusion, this is a book that everyone who wants to use jBPM5 to its full extend should get. I definitely recommend it for everyone who is interested in jBPM and Drools projects as it provides excellent introduction and much more than that.

Book is scheduled to be available in December 2012 so stay tuned.

Credits
Special thanks go to Maurico Salatino and Esteban Aliverti for this great book. Looking forward to the next one :)

2012/08/06

Simulation in jBPM (draft)

Recently some work has been started to provide simulation capabilities for jBPM. Simulation in many cases means different things to various people so let me start with context information on what simulation means to me and actually what current simulation component is capable of.

Simulation of business process is targeting business analysts that work on designing processes and optimizing them this is not a developer tool. It brings analitics into the picutre so while modelling the process various scenarios could be evaluated to see what is the best option based on current knowledge. In my eyes it is a way of learing the process and better understanding its design and prepare for the consequences a particular process can introduce. For example if we consider goods return process that will take care of products that were bought but for some reason was returned to the store. There are several steps that needs to be performed to analyze what is the reason of return:

it is broken
it does not meet customer expectations
etc

Depending on the reason various decisions could be taken, from rejecting the return, through sending that for further analysis to verify it is broken up to accepting it and returning funds. Following diagram illustrates sample of such process

We can imagine that this process is prepared for heavy sale period which could be Xmas time. Analyst considers that there could be quite some returns due to it was a duplicated/missed gift and is wondering how to prepare its company to deal with it efficiently. With simulation and just few additional information (input data) provided as some sort of forecast on the expected load and available resources (s)he could identify potential bottlenecks in the organization that will prevent it from gathering profit (working efficiently). To name few of these information:

probability of taking a given path on the gateway
time spent on executing given activity
how many people are available to work on user tasks

Once such information are provided a simulated run through this process is executed and result is gathered and presented to the analyst for inspection. Number of runs can be executed with various input data to exercise "what if" scenarios. Alternative sources of information can exist, for instance analyst can make use of real time data collected by business activity monitoring for processes that are already running on production systems.

With this short introduction we can move on to look into how this is realized by jbpm simulation.

First of all jbpm simulation is divided into two components:

path finder
simulation engine extension

path finder component is responsible for determining all alternative paths in the process to illustrate how a process can be traversed. This is not only informational but as well input for running simulation. Following image shows the sample process with identified alternative paths and one is visualized on the diagram.

simulation engine extension is (as name suggests) an extension to jbpm engine that allows you to run simulations instead of normal process instance executions. Instead of relying on process data like variable it will traverse the process based on identified paths. So that means that path finder component is responsible for providing input to the simulation engine extension component. This at least is the main use case,it could be used as well to alter the path flow in cases of debugging the simulation.

Simulation engine extension provides core of the simulation dedicated to processes but does not run the simulation itself. For that drools- simulator (some details about it can be found here and here) is employed together with its fluent api that is based on paths and steps that can be positioned in time.

A typical use case would look like this:

model process definition
determine all alternative paths
each alternative path will be a path in drools simulatoin fluent
define steps for the path (there could be several steps configured for a drools simulator path that in fact represents simulation instance) on a given time distance
add SimulateProcessPathCommand for each step
run simulation

To see a running example of such simulation take a look at test case that is part of the jbpm-simulation.

Simulation engine extension while executing simulation will generate events for every simulated activity those events will be stored in simulation repository that could have various capabilities. Personally I prefer one that is backed up with stateful knowledge session and can employ complex event processing and rules to provide meaningful simulation results.

This is just short heads up on the simulation efforts in jbpm so please leave your comments on what would you like to see supported by this component.

Further details about jbpm-simulation components can be found in jbpm wiki (soon).

2012/07/31

Service Task with web service implementation

Invocation of web services as part of business process is common and most likely because of that default implementation of Service Task in BPMN2 specification is web service. Recently (5.4) jBPM5 has gotten support for such activity.

jBPM5 web service support is based on Apache CXF dynamic client. It provides dedicated Service Task handler (that implements WorkItemHandler interface).

org.jbpm.process.workitem.bpmn2.ServiceTaskHandler

Worth noting is that this handler is capable of invoking both web service end points and simple Java based services as with previous ServiceTask handler (org.jbpm.bpmn2.handler.SendTaskHandler) based on the implementation attribute of service task node

web service implementation

 <bpmn2:serviceTask id="ServiceTask_1" name="Service Task" implementation="##WebService" operationRef="_2_ServiceOperation"> </bpmn2:serviceTask>

java implementation

<bpmn2:serviceTask id="_2" name="Hello" operationRef="_2_ServiceOperation" implementation="Other" >

</bpmn2:serviceTask>

ServiceTaskHandler can invoke web service operations in three modes:

synchronous (sends request and waits for response before continuing)
asynchronous (sends request and uses callback to get response)
one way (sends request and does not wait for any response)

This configuration is done on service node level as parameter (data input) so it allows to have different configuration for service nodes and to be handled by the same service task handler.

Let's try to go through this implementation with example. We are going to build a process that will get weather forecast for given zip codes in US. So process will look like this:

This process will:

ask for couple of zip codes on the first human task (task is assigned to john)
next transform result of the user task to a collection that will be used as input for multi instance service task
then based on the input collection process will create several instances of service task to query the weather forecast service
once all service task instances are completed it will log result to the console
and create another human task to show the weather forecast for selected zip codes (task is assigned to john)

When process instance is started it will prompt the user to select in what mode service task instances should be executed: async or sync. With this particular example changing the mode from async to sync will not make big difference as the service we use is rather fast but with services that takes some time to respond difference will be noticeable.

But how does it know it is web service and even more important what web service it is? This is configured as part of process definition using few dedicated constructs:

1. first of all we need to tell the engine where is our WSDL so it can be read and operations from it can be invoked - this is done with BPMN2 import:
<import importType="http://schemas.xmlsoap.org/wsdl/" location="http://wsf.cdyne.com/WeatherWS/Weather.asmx?WSDL" namespace="http://ws.cdyne.com/WeatherWS/"/>

2. next message, interface and operations must be defined:

<itemDefinition id="_2-2-4_InMessageType" />

<message id="_2-2-4_InMessage" itemRef="_2-2-4_InMessageType" />

<interface id="_2-2-4_ServiceInterface" name="" implementationRef="Weather">
 

  <operation id="_2-2-4_ServiceOperation"   

       implementationRef="GetCityWeatherByZIP" name="hello">
      <inMessageRef>_2-2-4_InMessage</inMessageRef>
  </operation>

</interface>

Important: make sure that implementationRef for both interface and operations point to valid service and operation in WSDL.

3. Next use defined operation in your service task and set implementation as web service (or don't specify that attribute at all so default will be taken):

<serviceTask id="_2" name="Service Task" operationRef="_2-2-4_ServiceOperation" implementation="##WebService" >

........ 

</serviceTask>

NOTE: Unfortunately tooling does not support this yet so the bpmn2 file needs to be edited by hand. Soon tooling will provide this as well.

Yet another important thing here is that if you plan to use request or response object of the service in your process as variables make sure that all of them implement java.io.Serializable interface so they can be properly persisted. One way to do this (used in the example) is to provide additional configuration to tell JAXB to add Serializable while generating classes from WSDL and generate classes as part of the build:

create extra binding file
configure cxf codegen maven plugin

Complete source code can be found here. It comes with test cases that uses this example as well as local service that can be used to illustrate difference between sync and async modes.

This example can be executed on jbpm-console when build from master, as it already has this service task handler configured. Here is a short guide on how to do it:
1. clone jbpm master and build it with mvn clean install -DskipTests -Dfull (alternatively download latest build from build server)
2. clone jbpm-examples and build jbpm-ws-example project with mvn clean install
3. copy result of build in point 2 (jbpm-ws-sample-1.0-SNAPSHOT.jar) into jbpm-installer/dependencies
4. copy WeatherWSServiceProcess.bpmn2 into jbpm-installer/sample/evaluation/src/main/resource
5. copy all archives from jbpm-distribution into jbpm-installer/lib
6. use jbpm-installer to install jbpm into jboss AS7 - ant install.demo.noeclipse
7. start demo server - ant start.demo.noeclipse

then go to jbpm-console and run the process with name WeatherWSServiceProcess.

Have fun!

2012/07/30

JUDCon2012 - Boston

JUDCon 2012 in Boston was a great place to meet with fellow developers and discuss about future of jBPM. I had a pleasure to give a talk about experimental project that I have been working on for quite some time, but unfortunately not enough time was spent on it. Anyway, I gave a introduction to the project that I use to call jbpm enterprise.

The goal of this project is to provide comprehensive BPM platform on top of jBPM and Drools projects to leverage the most of them in the enterprise (Java EE) environment. So what it does?

First of all it bundles jbpm and drools (together with all dependencies) into JBoss Module.
Next it provides very tiny layer that abstracts knowledge api and provides simplified interfaces to interact with execution engine (execution engine is knowledge base with sessions enclosed by an component with additional characteristics - more in presentation)
exposes some services via OSGi service repository
provides maven archetypes to build your components that utilize the platform over OSGi service registry

bundle archetype that is dedicated to put you process/rules/events logic in
web application archetype that makes use of the platform (starts, signal processes, etc)

More details can be found in the presentation from the JUDCon2012 conference.

jBPM&Drools go Enterprise - JUDCon2012

If some one would like to give it a test drive please find a downloadable artefacts and short guide here.

Have fun and as usual comments are welcome.

In this case your feedback is even more important to see if that is something that community expects.