rhpam Posts

How to Handle CORS in Red Hat Process Automation Manager

One feature on Red Hat Process Automation Manager (RHPAM) is the ability to provide an API endpoint which can be accessed from multiple applications. But when we call them directly from javascript within a browser, sometimes it would shows a CORS error.

Workaround is quite easy, either we add a CORS header on RHPAM API response or change call method from a browser call into server-to-server call. For this article i would go with the first approach, and that is adding a CORS header on API that RHPAM provides.

Luckily our RHPAM deployment is on top of Openshift, and by adding below key-value parameters to RHPAM kie-server’s Deployment Config is good enough to solve CORS issue.

         - name: FILTERS
           value: "Access-Control-Allow-Origin"
           value: "*"
           value: "Access-Control-Allow-Methods"
           value: "POST,GET,OPTIONS,PUT"
           value: "Access-Control-Allow-Headers"
           value: "*"
           value: "Access-Control-Allow-Credentials"
           value: "true"
           value: "Access-Control-Max-Age"
           value: "86400"

How to Fix Error “Variable content was trimmed as it was too long” on RHPAM

Just yesterday i got a very unique error,

Variable content was trimmed as it was too long (more than 255 characters)

I can see that this error happens due to RHPAM saves the content of each object on database, but have a 255 character as limitation. And for this case, my object content length is more than 255 characters.

Workaround for this is we need to alter database generated by RHPAM and for example below, im changing it from 255 into 5000 characters.

alter table VariableInstanceLog modify column value varchar(5000);

alter table VariableInstanceLog modify column oldValue varchar(5000);

And need to put below parameter runtime,


For a RHPAM deployment on Openshift, we can embed below parameters on RHPAM DeploymentConfig,

	  value: '-Dorg.jbpm.var.log.length=5000'

Restart RHPAM, and we can see now RHPAM can handle object with length up to 5000 characters.


Integrating DMN and Business Process on Red Hat Process Automation Manager

DMN stands for Decision Model and Notation. According to Wikipedia it is a standard approach for describing and modeling repeatable decisions within organizations to ensure that decision models are interchangeable across organizations. It is another approach of creating a “decision” on RHPAM (Red Hat Process Automation Manager), other than Decision Table and DRL.

On RHPAM, DMN file can be deployed as a standalone dpeloyment, or as an embedded within a Business Process. On this writing, im trying to do both and we’ll see what are the benefit and weakness of each approach.

Lets try to create a simple DMN to calculate how much loan should one get based on his age and salary. Create “age” and “salary” as DMN Input Data, and two DMN Decisions. “Loan_limit” with Decision Table, and “result” with Context.

Save, Build and Deploy, and we can test it by using rest api. But first we need to check on DMN Namespace and Model Name which is highlighted on the first screenshot and then put is as json parameter.

curl -L -X POST 'http://localhost:8080/kie-server/services/rest/server/containers/loan_validation_1.0.0-SNAPSHOT/dmn' \
-H 'Authorization: Basic cGFtQWRtaW46cGFzc3dvcmQ=' \
-H 'Content-Type: application/json' \
--data-raw '{
  "dmn-context": {

The next step is put this DMN into a workflow. We can start by crating a simple workflow, dont forget to add DMN Namespace and Model Name on Business Rule Task.

And run this curl ommand to create a new instance,

curl -L -X POST 'http://localhost:8080/kie-server/services/rest/server/containers/loan_validation_1.0.0-SNAPSHOT/processes/loan_workflow/instances/' \
-H 'Authorization: Basic cGFtQWRtaW46cGFzc3dvcmQ=' \
-H 'Content-Type: application/json' \
--data-raw '{

And we can see the result on log,

12:58:08,486 INFO  [stdout] (default task-17) ================
12:58:08,486 INFO  [stdout] (default task-17) you are eligible for 20000
12:58:08,486 INFO  [stdout] (default task-17) ================

Code sample can be downloaded on below github repository,


A Simple RHPAM-Application Integration with Multi User Approval Workflow

On this article, im trying to create a simple Leave Approval System which is deployed as microservices. So basically there are two different application, one Golang app as Frontend, and RHPAM (Red Hat Process Automation Manager) as Backend, and both are communicating by using a simple REST API.

Basically there are two users involved here, one is adminUser as requester, and spv02 as approver. As approver, spv02 can either accept or reject the requested leave. Easiest way to described the approval workflow is perhaps described in below image,

Lets start by creating two user on RHPAM,

adminUser / password
spv02 / password

Import project from github (code is at the end of this article), build and deployed it to KIE.

Now lets start with creating an entry form,

The touch-point between frontend and pam on this ui is RHPAM’s “process instances” API, and im adding a correlation-key which is a unique business primary key.

curl -L -X POST 'http://localhost:8080/kie-server/services/rest/server/containers/approval_system_1.0.1-SNAPSHOT/processes/approval/instances/correlation/TL-3422' \
-H 'Authorization: Basic YWRtaW5Vc2VyOnBhc3N3b3Jk' \
-H 'Content-Type: application/json' \
--data-raw '{
    "application": {
        "com.myspace.approval_system.Request": {
            "days": 9,
            "purpose":"Sick Leave"

Next step is displaying all leave request that have been made by this corresponding user, we can capture this by using server queries API, given a specific username as initiator parameter,

curl -L -X GET 'http://localhost:8080/kie-server/services/rest/server/queries/processes/instances?initiator=adminUser&page=0&pageSize=10&sortOrder=true&status=1&status=2&status=3' \
-H 'Accept: application/json' \
-H 'Authorization: Basic YWRtaW5Vc2VyOnBhc3N3b3Jk'

Moving forward, now we are seeing from approval’s point of view, first we need to display what are the tasks which are assign to this user.

curl -L -X GET 'http://localhost:8080/kie-server/services/rest/server/queries/tasks/instances/owners?page=0&pageSize=10&sortOrder=true&sort=taskId' \
-H 'Accept: application/json' \
-H 'Authorization: Basic c3B2MDI6cGFzc3dvcmQ='

And the ability to Approve or Reject a specific request, there are two APIs which are involved here. That is one API for starting the task, and another one to complete it. Make sure you differentiate parameters on “approved” field, use “false” for Rejecting request, and “true” for Accepting request.

curl -L -X PUT 'http://localhost:8080/kie-server/services/rest/server/containers/approval_system_1.0.1-SNAPSHOT/tasks/6/states/started' \
-H 'Authorization: Basic c3B2MDI6cGFzc3dvcmQ='
curl -L -X PUT 'http://localhost:8080/kie-server/services/rest/server/containers/approval_system_1.0.1-SNAPSHOT/tasks/6/states/completed' \
-H 'Authorization: Basic c3B2MDI6cGFzc3dvcmQ=' \
-H 'Content-Type: application/json' \
--data-raw '{
    "approved": false

And here are my repositories on Github,

frontend : 

backend :

Have fun playing with RHPAM :)


Reusing Your Workflow and Assets on Red Hat Process Automation Manager

Sometimes you want to reuse the assets which you are already creating, and “import” it on multiple other projects. We can use that on RHPAM (or its opensource version, which is jBPM) because the output of it is actually a kjar (knowledge jar) file, which is just a jar library where you can import it by using a simple maven.

For this example im creating two different project, Project 01 and Project 02. Where Project 01 have a workflow that will import a Decision Table which is created on Project 02. It is basically a simple calculator to measure how wealthy someone is, based on one’s salary.

Lets start with Project02 first, and as always we start with a simple Java bean

And a Decision Table,

We will use that decision table in a very simple workflow,

Next is the unique part, where we need to create a KIE base for Project02

And a KIE session,

“Default” option should be un-check for KIE bases and KIE session on Project02.

The last step for Project02 is to build, and install it.

Next is how to import Project02 into Project01. First thing that we need to do is to import Project02 on Project01’s dependency.

Create a workflow, and put a “Reusable Sub-Process” referring to our Project02 workflow.

And finally, create a KIE base for Project01

and a KIE session,

Make sure you check “Default” option for KIE bases and KIE session on Project01.

Build, Install and Deploy Project01 to KIE server, and we can test by using a simple CURL REST API command or by using Postman

For complete code, can be accessed here.