Red Hat
Jul 6, 2016
by Christina Lin
This year in Red Hat summit, I have to honor to work with two intelligent colleague of mine Eric Schabell and Siamak Sadeghianfar to share an interesting workshop. The main theme on the workshop is demonstrate how rules automation can be implemented on the cloud and how to take advantage of the integration service to expose API and scalable across the network.

Red Hat OpenShift, is a Platform-as-a-Service (PaaS) that allows developers to quickly develop, host, and scale applications in a cloud environment.  The lab goes through basics concept of OpenShift, explaining the following terminology in OpenShift.

Container: Lightweight mechanisms for isolating running processes
Image: A binary that includes all of the requirements for running a single container, as well as metadata describing its needs and capabilities.
Pod: A small group of tightly coupled Containers sharing network, storage, etc
Service: A set of replicated pods that work together
Build: Process of transforming a set of parameters and source code into a runnable image
Route: A route is a way to expose a service by giving it an externally-reachable hostname like www.example.com
Template: Describes a set of objects that can be parameterized and processed to produce a list of objects for creation by OpenShift

As always, I want to focus more on the software development side, lets take a look at what is going on in the lab. The background of the scenario is a weight watcher program, which allows users to register with their weight loss target. Once registered, it will base on daily input from devices sent in to the system to provide calculated statistic of how this person is doing. 

Red Hat  JBoss BRMS, using the declarative logic with business rules to build applications with independent modularized rulesets. In our lab, we will deploy the business rule to a separate decision server. Here we use it store all the calculation logic of how your weight has progressed and to diagnose how the person is doing. A decision server will be deployed on the cloud and a project of set of rules will be applied to the server. Once the rules are loaded the server become our rule engine and ready to take in user data via web service. 

Having all the rules set up and start watching the weight providing advices is good. But we are now in the era of APIs, we want to maximize the use of the rules, creating a simple RESTFul API, it offers benefits such as additional decoupling, and allows extreme scalability. Since REST forbids conversational state, we can scale very wide by utilizing load balancer later on provided automatically by OpenShift. Other advantage such as uniform interface with Basic API operations makes other parties easy to follow. Red Hat JBoss Fuse is the perfect medium to turn the Business Rules into simple services API, it’s built-in interface. As it's REST DSL allows user to expose web API in simple context as well as have 150+ variety supported component to connect other service available. 


This example expose two RESTFul Endpoints, one takes in participants data and one receives constant updates of weight and returns the analyzed result back. There are few things we need to do 
  • Create an event base system by exposing RESTFul Endpoint using REST DSL with Camel 
  • Transform the entered data into JSON format required by BRMS, and then wrap these data in XML which is needed by decision server. 
  • Connecting to BRMS decision server with API Call 
  • Filter and parse the return result 
I have created demo that will automatically setup both BRMS and Fuse on OpenShift, 

       https://github.com/jbossdemocentral/jboss-fis-weightwatcher

The install and running instruction are as follows, 

Installation

Make sure you have installed Vagrant and Virtual box. run initCDK.sh to setup CDK and web services (SKIP this section if you have already installed)
$ ./initCDK.sh
Install the project on OpenShift by running init.sh
$ ./init.sh
Login to OpenShift console with USERNAME/PWD admin/admin
https://10.1.2.2:8443/console/                                    
Overview BrokerOverview TCP BrokerOverview FUSE BRMS
Also Pods, services and route
Overview BrokerOverview TCP BrokerOverview FUSE BRMS

Running the demo

Registering a new Participants: Before we start sending in weight, we need to register the participants and set their goal.
http://weightwatcherapi-fisdemo.rhel-cdk.10.1.2.2.xip.io/weightwatcher/addParticipant/2/ipavlov@behaviorist.org/75/84/70/78
Add Participant
Events - Updating weight: In a browser or any REST client app, trigger the event by updating the weight as we measure every-time. Each event will then pass into rule engine through our API. You should be able to see the calculated response from weight watcher.
http://weightwatcherapi-fisdemo.rhel-cdk.10.1.2.2.xip.io/weightwatcher/updateweight/2/2015-04-20%2007:15:00%20EST/79

http://weightwatcherapi-fisdemo.rhel-cdk.10.1.2.2.xip.io/weightwatcher/updateweight/2/2015-04-25%2007:15:00%20EST/77

http://weightwatcherapi-fisdemo.rhel-cdk.10.1.2.2.xip.io/weightwatcher/updateweight/2/2015-04-28%2007:15:00%20EST/76
Update weightCamel status




Original Post