best practices in deploying ibm operation decision manager standard 8.8.0

Best Practices in Deploying IBM Operational Decision Manager StandardBDM-4361

Pierre Feillet - IBM ODM Product Architect

Agenda

• ODM Standard Architecture & runtime elements• Recommended topologies• Configuration• Decision Service Development Lifecycle• Business Rules everywhere• Decision Services in Apache Spark• Performance• Best practices

2

ODM Standard 8.8.0 Architecture

ODM 8.8.0 Family

• Standard• Focused on Decision

Services• Decision Center• Decision Server Rules

• Advanced = Standard +• Decision Server Insights

• Dynamic modeling to define Events, Entities & detection rules to search for situations

• Server Grid with events & data in memory

• WAS Liberty + WXS

4

IBM ODM Advanced

IBM ODM Standard

Decision Server AdvancedDecision Server RulesDecision Server Events

Decision Server Insights

Decision Server StandardDecision Server Rules

Decision Center

Decision Center

ODM Standard - Concept of Operations

5

Decision Server

Development

Operation

Continuous Improvement

Rule ExecutionServer

JEE/J2SE

Rule Designer(Eclipse)

Decision Center -Decision Validation Services

z Rule ExecutionServer

COBOL

Decision Center Repository

IT Architect

IT Developer

Business Analyst

Process Owner

Business UserBusiness Leader

Code Generation(deprecated)

Deploy

Synchronize

ODM Standard Component Organization

6

• Rule Designer• Decision Center

• Business Console• Enterprise Console

•Decision Server Rules• Rule Execution Server

• Rule Execution Server web console

• Hosted Transparent Decision Service

• Scenario Service Provider• Decision Warehouse• Rule Sessions• eXecution Unit

ODM Standard Component List

7

Component Acronym Description DB tables SIBus

Decision Center DC Web consoles + the DM authoring repository to store rule and event projects

Decision Center tables

None

Decision Server / Rule Server Console

RES Console

Executable Decision Service mgt through RuleApp/ruleset assets

Rule Execution Server tables

None

Decision Server / Rule eXecution Unit

RES XU Rule server core runtime deployed as JCA or Pojo Rule Execution Server tables

None

Decision Server / Rule Sessions

Rule Sessions

Rule runtime components available as EJBs or Pojo None ODM Bus

Decision Server / Decision Services

Custom or OOTB JEE artifacts that encompass the invocation and the decision logic (RuleApp/Ruleset). Are based on Rule Sessions.

None None expected unless MDB used

Decision Server / Scenario Service Provider

SSP Test and simulation rule runtime None None

Decision Server / HTDS HTDS Dynamic Web Services None None

Decision Server / Decision Warehouse

DW Runtime that listens and store decisions Decision Warehouse tables

None

A full fledge Enterprise Proposal

8

System z Enterprise 196 (or 114) +

LPAR, Sysplex

OS Distributed including z/Linux z/OS

100% Java product on distributed plus a native secret sauce in zRES

Power or Intel blades

z/VM z/OS

z/Linux

WAS

ND & L WAS

NDCICS

ODM ODM

distributed

WAS ND Other AS

Works with a variety of stacks WAS ND, WAS Liberty profile (8.5.5.x) on distributed & z/OS

zRES Lightweight server for z/OS, run standalone and in CICS

JSE Rule Execution Server & rule engine

ODM

ODM Technical Detailed System Requirements

Distributed

WAS

Liberty

WAS

Liberty

https://www-304.ibm.com/support/docview.wss?uid=swg27023067

https://www-304.ibm.com/support/docview.wss?uid=swg27023067

ODM Standard Consoles & Touch Points

9

Decision Center

Decision Server

ODM cell

Rule Designer

Decision Center Consoles

Desktop

WAS Administration Console

Rule Execution Server Console

DC Ant Tasks

DS Ant Tasks

Synchro over HTTP(S)

Executable artifact deploymentover HTTP(S)

All Consoles, synchronization, deployment going through HTTP(S)

HTTP(S)

HTTP(S)

HTTP(S)

HTDS WS endpoints

HTTP(S)

Zoom on Decision Server / Rule Architecture

10

RuleApp & Ruleset archive deployment

Rule Sessions EJBPOJO

EngineJCA

CustomDecision Services

eXecution Unit

Management Console

RuleApp/Ruleset Repository

EngineEngine

Mgt Notification

JMX/IP

Management Model

Decision Warehouse DB Decision Traces

Scenarion Service Provider

JMX

10

HostedTransparent

DecisionServices

WS WS, JSE, Message, your choice

RES Console

• Workload– Only management of the rule executable artifacts (RuleApp, ruleset,

eXecutable Object Models)– Execution restricted to diagnostic and interactive testing– Notification to eXecution Units when a new artifact comes available

• Design– In memory stateful webapp on the top of the RES DB– Contains a JMX Model

• Kind of WAS Deployment Manager for Decision Services

11

RES Console Out of a Cluster

• A singleton in a WAS Cell• Deployed apart a cluster in reference topology

– Install the RES Console on a separated server– SPOF but with limited impact if it fails

12

RES What If Scenarios

• What if RES Console shuts down– What do we loose

• Deployment & notification of new ruleapps/rulesets versions through HTTP(S)• REST Mgt API become unavailable• Execution statistics

– What do we keep• Rule execution remains fully operational• Direct deployment to the RES DB remains possible

– Restart RES Console to dynamically get back these functionalities• What if RES DB goes down

– Execution remains fully operational for all rulesets already in memory

13

RES Notification Protocol

• JMX• Always active• Works in a WAS Cell or equivalent giving a JMX MBean server federation• RES Console defines its RES model top level MBean• Each XU defines its Mbean

• IP• Set in RES Console and Xus descriptors• RES Console hostname/IP & port number specified in each eXecution Unit to get

notified• Rely on a virtual address (VRRP) to keep XUs untouched when RES Console is

moved of machine or IP address• Enter a RFE if you are interested in enlisting a set of RES Console IP addresses

14

ODM Standard TopologiesReference Topologies

How to choose your ODM topology

16

What is my Decision Management scope?

Decision servicesSituation detection

Dev phase

Dev Integration Testing ProductionHA/DR

What is the workload nature ?

Event processing Remote/Local decision execution

Business decision Validation through Simulation and Testing

What is the platform?

distributed z: z/OS, z/Linux Hybrid Cloud

User Testing PreProd

ODM Standard Topology cheat sheet• Abbreviations for product scopes

– ODM: Full platform

– ODMR: Rules only

– ODME: Event only

– ODM/DC: Decision Center

– ODM/DS: Decision Server

– ODM/DSR:DS for Rules only

• WAS ND topology concepts• ODM Profile templates delivered for clustered DC &

DS

17

Cluster

DMGRNode Agent

Server

Server

Node Agent

Server

Server

Dmgr’s nodeCustom nodeCustom node

Cell

ODM Standard Gold topology: DC + DS clusters + RES Console

• Functional separation/isolation of workload

• Several cells for staged dev lifecycle

• RES XU deployed at Node level

• Datasources deployed at cluster level

• Remote ClientApp cluster run in another cell or can be added to the ODM cell

• Can scale individual clusters as needed

Cell

Node Agent

Cluster Member 1HTDS for execSSP for testing

Rule Session jar if Custom Decision

Services

Cluster Member 1Decision Center


DecisionCenterCluster

Decision ServerCluster



ServicesDecision Center DB

Decision Server Rule

DB

Decision Warehouse

DB

IP Sprayer

RES XU RAR

Machine 1 Machine 2

Server 1RES Console

Cluster Member NDecision Center

Cluster Member NHTDS for execSSP for testing


Services

Machine N

Deployment Manager

IHS IHS IHS

Node Agent

RES XU RAR

Node Agent

RES XU RAR

ODM: DC + DS + RES Console + Remote Client App clusters


• Several cells for staged dev lifecycle

• RES XU deployed at Node level

• Datasources deployed at cluster level

• SSPs are directly referenced on each member from DC

Cell

Node Agent



Services







Services

Decision Center DB


DB

Decision Warehouse

DB

IP Sprayer

RES XU RAR

Machine 1 Machine 2

Server 1RES Console




Services

Machine N

Deployment Manager

IHS IHS IHS

Node Agent

RES XU RAR

Node Agent

RES XU RAR

Cluster Member 1Client App

Cluster Member 3Client App

ClientApp

ClusterCluster Member N

Client App

ODM Standard Challenger Topology: DC cluster + DS clusters + RES Console Started/Stopped cluster

• Variant of the reference topology

• Used by few customers

• *Should* work in a general approach but not yet fully tested in the Lab

Cell

Node Agent



Services







ServicesDecision Center DB


DB

Decision Warehouse

DB

IP Sprayer

RES XU RAR

Machine 1 Machine 2

Server 1RES Console




Services

Machine N

Deployment Manager

IHS IHS IHS

Node Agent

RES XU RAR

Node Agent

RES XU RAR

RES ConsoleCluster

Server 2RES Console

(stopped)

Server 3RES Console

(stopped)

ODM: Testing & Simulation cluster


• SSPs can be referred by their cluster address when using Testing & Simulation from the Business Console (CDI)

• SSPs are directly referenced on each member from DC when using Enterprise Console

Cell

Node Agent

Cluster Member 1SSP

DecisionSimulation

ClusterCluster Member 2

SSP


DB

Decision Warehouse

DB

IP Sprayer

RES XU RAR

Machine 1 Machine 2

Server 1RES Console

Cluster Member NSSP

Machine N

Deployment Manager

IHS IHS IHS

Node Agent

RES XU RAR

Node Agent

RES XU RAR

ODM Standard: Standalone Server

Full ODM Standard running in a single server

Includes all runtime parts for web decision authoring, test and execution

Server1Decision Center

RES Console HTDS for exec

SSP for testing

Rule Session jar if Custom

Decision ServicesXU RAR


DB

Decision Warehouse

DB Machine 1

Decision Center DB

ODM Standard: Standalone Server

Decision Center on single server + Decision Server single server

2 servers environments: • no failover• 2 minimal # of JVMs• easy to create (create standalone profile) • Automatic DB creation

Server1Decision Center


DB

Decision Warehouse

DBMachine 1

Decision Center DB

Server1RES Console HTDS for exec

SSP for testing

Rule Session jar if Custom

Decision ServicesXU RAR

Machine 1

Operational Decision Management Scope

Functional restriction

Project Phase Recommended Topology

ODM Standardmeaning DC + DS for Rules

Dev ODM - standalone

Test ODM – from 2 servers to DC + DS cluster (Gold)

Prod ODM DC + DS clusters (Gold)

ODM / Decision Center Dev ODM/DC standalone

Test ODM/DC 1 cluster

Prod ODM/DC 1 cluster

ODM / Decision Server Dev ODM/DS – standalone

Test ODM/DS - 1 cluster

Prod ODM/DS – 1 to 4 clusters

ODM/Decision Server for Rule Rule Dev ODM/DSR – standalone

Test ODM/DSR – 1 cluster

Prod ODM/DSR – 1 or more clusters

Rule Simulation Dev ODM/DSR/TS – standalone

Rule Simulation Prod ODM/DSR /TS– 1 cluster

ODM Standard Topology Catalog

ODM CopperDC + DS combined in a standalone server

ODM Bronze1 cell with 1 DC server + 1 DS Server

ODM Silver1 cell with a single cluster for DC & DS

ODM Gold topology1 cell with a DC cluster & DS clusterSeparation between business authoring from executionCover all product usage including business authoring, execution and simulation for rule and event Best tradeoff between Performance, HA, and cost

Platinum topologies All upper topologies that provide more workload separation and usage specialization Client App cluster can be added in this cell or run in a separated cellSimulation is isolated from business authoring & pure execution

ODM Topologies at a glance

Components & ConfigurationSubhead

Decision Server endpoints declared in Decision Center Enterprise ConsoleExecution, Test, or both

If you use Testing & Simulation from Decision Center – Enterprise ConsoleDeclare SSP cluster members in Decision CenterSSP.EAR/SSP.WAR does not support to be load balanced

If you use Testing & Simulation from Decision Center – Business ConsoleDeclare the cluster url for SSPs

An IlrSSPResolver implementation may be needed in DC when using T&S from Entreprise Console & Rule Designer

Decision Server Definitions in Decision Center

Backend application is ssp.ear for scenario service providerComposed of 2 web apps

ssp.warUsed to run Test Suites and Simulations from Decision Center Enterprise Console and Rule DesignerDeployment of ruleset always performed via local API & JMX or through HTTP(S)SSP.EAR/SSP.WAR/web.xml LOCAL_DEPLOYMENT_ONLY property set by default to false

Decisionrunner.war – used to run Test Suites and Simulations from BC/Business Console

Deployment of ruleset always performed via local API & JMXDecision Service Project versus Rule Project

SSP Configuration for RES Deployment

Decision Service Life CycleSubhead

Decision Management Development Life Cycle

In a strict isolation required between types/purposes of environments:• DEV: application developers have access to “DEV” only , purpose is application dev and basic function testing• SIT: integration testers have access to “SIT” only, purpose is integration and function testing• User Acceptance Test: user acceptance testers have access to “UAT” only, purpose is user testing• PREPROD: function and integration proven in earlier environments, purpose of this env is largely perf testing• PROD - apps were proven in earlier environments. Barely anyone has access to this env, it is tightly regulated.

DEV SIT UAT PREPROD PROD

DEV SIT UAT PreProd

PROD

Decision Management Life Cycle Features are deployed in phased environment depending on purpose Business authoring

Business Object Model and Rule Editing Business Simulation & Test

Create and run Business Test Suites and Simulation to guide and improve decision logicChampion/Challenger comparison on KPI computation

ExecutionDecision run

Business Authoring

ExecutionBusiness Simulation & Testing

DEV SIT UAT PreProd PROD

Classic Decision Management Life Cycle Development of delimited Decision Services Business Authoring stopping most likely at UAT Business Simulation used until PreProd max

Business Authoring

ExecutionBusiness Simulation & Testing

Agile Decision Management Life Cycle

Agile ODM Provide an ODM as a PaaS with change/extension of the decision logic

without starting back from Dev environment LOB leverage business authoring, simulation and apply their corporate

decision logic on their data Business Authoring and Simulation available in PreProd and even Prod More agility coming with more risk in production

Need governance and workload isolation to handle business authoring, simulation and execution and avoid regressions

DEV SIT UAT PreProd PROD

Execution

Business Simulation

Business Authoring

ODM Deployment across phases Each life cycle phase is mapped into an environment

Each environment is a WAS cell (or equivalent in other AS) Decision Server deployed in isolation by phase Decision Center can be shared or isolated per phase

1 or more cells dedicated by phasePhase environment contain 1 DC (cluster) when business authoring required or

1 unique shared Cell with Decision Center + multiple Cells for Decision Server

This configuration is the most commonly usedMultiple DC are possible for sandbox

Shared DC across environments• One Decision Center managing multiple Decision Servers• DC clustered with a HA/DR DB• User access, Branches mgt for multi team/releases activities• DC and DS can be split into 2 isolated envs• Online and offline RuleApp/Ruleset deployment supported in DS

Decision Server

Decision Server

DS env

Dev Integration Testing

Executable Decision Service deployment over HTTP(S)

Decision Server

DS env

Decision Server

DS env

UAT PreProd

Decision Server

DS env

Production

DS env

Decision Center

DC env

SDLC – Shared Decision Center

One source of truth for business rule authoring Deploy on all Decision Servers; Multiple DS/RES are supported by DC HA with DC and DS in cluster Leverage project branch and merge in DC and deploy executable rules on the various DS Access to projects and DS scoped by groups & users in DC Isolation is preserved for the execution runtime

Need for other DC installation for upgrade and DR Applied by most customers

Leverage clustering and HA/DR DB

Best practice is to minimize the number of DCs and have 1 source of truth

Business Rules everywhereAuthor once execute everywhere

zRule Execution Server Stand-alone WebSphere Application

Server for z/OS

WOLA

CICS

COBOL Application

Rule Execution Server for WAS for z/OS

COBOL <-> JavaMarshaller

COBOL Generation

Rules

GeneratedCOBOL

JVM ServerzRES

zRule Execution

Server

IMS

COBOL Application

z/OS Batch

COBOL Application

COBOL Generation

Rules

GeneratedCOBOL

COBOL Generation

Rules

GeneratedCOBOL

DS StubDS Stub DS Stub

Decision invocations flavors in z/OS

1 2

3

1

32

zRES

RES on WAS on z/OS

COBOL gen

3 options

ODM in the Clouds

• ODM on Cloud– Standard edition in SaaS

• ODM Patterns– in PureAS (client IT)– In Pure Service (SoftLayer)

• Bluemix Business Rules– A subset of ODM ready to use in Pay as you Go approach– Deploy your decision services in a PaaS– Based on Cloud Foundry– RES Console + HTDS + Rule Designer update site

• Support on Hypervisors for an installation on IaaS

Decision Services in Apache SparkBatch & more if affinities

Spark

• Decision Services running in Apache Spark based• Opportunity to execute large batches but also

– Post-process metrics & KPIs for simulation and production– Integrate with Machine Learning and analytics algorithms available in Spark

• RES Java APIs are callable from Scala• Decision Server Rules fully compatible with Spark• Article: https://developer.ibm.com/odm/docs/solutions/odm-and-

analytics/odm-business-rules-with-apache-spark-batch-operations/

Running a Decision Service in a Spark cluster

42

Best PracticesRecommendations in deploying rule based decision services

Best Practices

• Prefer 1 shared DC + multiple DS dedicated per env– Use the branch/merge for handle team collaboration & multiple releases– Additional DCs are possible but adding artifact lifecycle mgt complexity

• 1 WAS Cell per stage env with Decision Server• Leverage AS cluster for preprod and production envs• Add DB HA/DR for DC and DS production envs• Use collocated DC+DS for standalone dev envs• Make a dedicated simulation env when this workload become

significant

44

Best Practices

• Automate RES Console restart• Choose your RES notif in Decision Server

– JMX notification when RES Console & all eXecution Units are in the same AS admin scope (WAS cell or equiv)

– IP notif when running on JEE/JSE without JMX federation

45

Wrap up

• ODM Standard supports a variety of deployments on cloud on prem and hybrid

• Adopt latest features to get the best of the product: Decision Service projects, Decision Engine, Business Console, Managed XOM in DC

• Tell us if you are interested in new platforms & languages: embedded, mobile, Javascript in web browser

• Help the Lab by entering RFEs to push for a full RES Console cluster support

• For more– ODM topology DevWorks article– ODM Performance Redbook

46

Notices and Disclaimers

47

Copyright © 2016 by International Business Machines Corporation (IBM). No part of this document may be reproduced or transmitted in any form without written permission from IBM.

U.S. Government Users Restricted Rights - Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM.

Information in these presentations (including information relating to products that have not yet been announced by IBM) has been reviewed for accuracy as of the date of initial publication and could include unintentional technical or typographical errors. IBM shall have no responsibility to update this information. THIS DOCUMENT IS DISTRIBUTED "AS IS" WITHOUT ANY WARRANTY, EITHER EXPRESS OR IMPLIED. IN NO EVENT SHALL IBM BE LIABLE FOR ANY DAMAGE ARISING FROM THE USE OF THIS INFORMATION, INCLUDING BUT NOT LIMITED TO, LOSS OF DATA, BUSINESS INTERRUPTION, LOSS OF PROFIT OR LOSS OF OPPORTUNITY. IBM products and services are warranted according to the terms and conditions of the agreements under which they are provided.

Any statements regarding IBM's future direction, intent or product plans are subject to change or withdrawal without notice.

Performance data contained herein was generally obtained in a controlled, isolated environments. Customer examples are presented as illustrations of how those customers have used IBM products and the results they may have achieved. Actual performance, cost, savings or other results in other operating environments may vary.

References in this document to IBM products, programs, or services does not imply that IBM intends to make such products, programs or services available in all countries in which IBM operates or does business.

Workshops, sessions and associated materials may have been prepared by independent session speakers, and do not necessarily reflect the views of IBM. All materials and discussions are provided for informational purposes only, and are neither intended to, nor shall constitute legal or other guidance or advice to any individual participant or their specific situation.

It is the customer’s responsibility to insure its own compliance with legal requirements and to obtain advice of competent legal counsel as to the identification and interpretation of any relevant laws and regulatory requirements that may affect the customer’s business and any actions the customer may need to take to comply with such laws. IBM does not provide legal advice or represent or warrant that its services or products will ensure that the customer is in compliance with any law

Notices and Disclaimers Con’t.

48

Information concerning non-IBM products was obtained from the suppliers of those products, their published announcements or other publicly available sources. IBM has not tested those products in connection with this publication and cannot confirm the accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the capabilities of non-IBM products should be addressed to the suppliers of those products. IBM does not warrant the quality of any third-party products, or the ability of any such third-party products to interoperate with IBM’s products. IBM EXPRESSLY DISCLAIMS ALL WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.

The provision of the information contained h erein is not intended to, and does not, grant any right or license under any IBM patents, copyrights, trademarks or other intellectual property right.

IBM, the IBM logo, ibm.com, Aspera®, Bluemix, Blueworks Live, CICS, Clearcase, Cognos®, DOORS®, Emptoris®, Enterprise Document Management System™, FASP®, FileNet®, Global Business Services ®, Global Technology Services ®, IBM ExperienceOne™, IBM SmartCloud®, IBM Social Business®, Information on Demand, ILOG, Maximo®, MQIntegrator®, MQSeries®, Netcool®, OMEGAMON, OpenPower, PureAnalytics™, PureApplication®, pureCluster™, PureCoverage®, PureData®, PureExperience®, PureFlex®, pureQuery®, pureScale®, PureSystems®, QRadar®, Rational®, Rhapsody®, Smarter Commerce®, SoDA, SPSS, Sterling Commerce®, StoredIQ, Tealeaf®, Tivoli®, Trusteer®, Unica®, urban{code}®, Watson, WebSphere®, Worklight®, X-Force® and System z® Z/OS, are trademarks of International Business Machines Corporation, registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the Web at "Copyright and trademark information" at: www.ibm.com/legal/copytrade.shtml.

http://www.ibm.com/legal/copytrade.shtml

Thank YouYour Feedback is Important!

Access the InterConnect 2016 Conference Attendee Portal to complete your session surveys from your

smartphone, laptop or conference kiosk.