TIBCO Collaborative Information Manager Architecture

The TIBCO Collaborative Information Manager application is the foundation of TIBCO’s MDM solution. TIBCO CIM delivers the key components required to achieve comprehensive MDM.

• Core application components

 - Information repository

 - Data governance

 - Process Automation

 - Internal and external synchronization

 - Reporting and business intelligence

• Support for multiple strategic data domains

(product, customer, vendor, location, financial data, reference data, code tables, etc.)

• Vertical templates

LOOSELY COUPLED PRODUCT ARCHITECTURE

The TIBCO CIM application is designed in a loosely coupled manner with well defined standards-based interfaces between its internal components. Process templates are built on top of a flexible data model that relies on the five core application components to power the solution. Abstracting these layers allows for a highly flexible and scalable offering that can be quickly configured to meet the customer’s requirements. Furthermore it provides the skill to support multiple data domains (product, customer, vendor, etc.) from one platform.

GUI

CIM is an entirely thin client application except for a thick-client optional graphical process modeler. The web user interface is totally metadata driven so screens are dynamic and contextual. Information is passed back and forth between the engine and UI components as XML; hence it is easy to swap out TIBCO CIM screens and embed custom rich-client interfaces right into the application using tools such as TIBCO’s Ajax-based product TIBCO General Interface.

The CIM GUI is based on the user interface design practices as laid out by the TIBCO One initiative, a corporate wide proposal to make available customers with a best in-class unified UI experience across products.

The CIM GUI uses web 2.0 technologies such as AJAX throughout the application for graphically creating input maps, installation and configuration, building relationships, etc.

High Performance

TIBCO CIM makes sharp use of distributed caching and parallelization of work across multiple threads to load, process, and publish large amounts of information very quickly. This is essential in data-intensive industries such as economic services, telecom, and energy. TIBCO CIM can automatically split large record sets into smaller batches for asynchronous dispensation with parallel threads working on individual batches. This greatly enhances performance and allows for the use of all available hardware by the system. Batch sizes can be configured via a property file. Inserts into the repository also occur in larger batch sizes rather than as individual Transactions.  also, TIBCO CIM can be configured to load record data into the cache at startup. The validation and process engines would then be retrieving information from the cache and could evaluate the rule base and load the data into the repository considerably faster than before.

TIBCO CIM aggressively uses the cache to speed performance by caching workflow states, Web service responses, user login information, temporary objects, etc. By gradually moving to a cache-first architecture with intelligent refresh and write-behind CIM leads the market with respect to performance and high volume deployments.

The TIBCO professional service group provides best practice guidance on making optimal performance decisions such as structuring the data model and relationships, incrementing activities in the workflow, and executing activities in a synchronous or asynchronous manner

WEB SERVICES

Application functionality – down to a fairly granular level such as querying and updating the structure of the master catalog, adding or modifying records, firing a business process, or querying and submitting work items within a business process, to name a few – are all exposed via pre-packaged web services. This makes it easy for customers who wish to use their own custom UI external the application to call TIBCO CIM services or embed TIBCO CIM services as part of a larger composite application. Both functional and highly granular web services are provided out of the box, so customers have a great deal of flexibility in accessing and controlling TIBCO CIM functionality from outside the application. This yields a master data services platform to maintain SOA and BPM initiatives that need reliable access to accurate and consistent data.

Application-Based: TIBCO CIM is a self-contained MDM application that provides a tremendous amount of out-of-the-box functionality, counting data models, rules, and business processes. TIBCO CIM has a wizard-driven installation and configuration process and does not need any other TIBCO software and can work with other vendor’s integration and business-to-business platforms as necessary.

Flexibility: By embedding the rules engine, process engine, and synchronization engine as part of the application, these engines are “aware” of the data model. This means that any changes to the data model such as adding or modifying an attribute are instantly available to the rules and process engine and instantly available for downstream consumption without the need to re-code any interfaces. Further, changes to the data model do not disrupt the system and can be hot deployed. This versatility differentiates the application from the competition and lower the cost of ownership.

Configuration vs. Customization: A solution that can be configured rather than built at the client site leads to faster deployments and a lower TCO. This includes configuring the data model via the GUI, importing existing metadata via the GUI, configuring rules either via the GUI or through the XML source files, and configuring processes via a graphical process modeler without the need to write code. The configuration vs. customization approach leads to seamless upgrades, lower services costs, and higher project success rates.

STANDARDS : TIBCO CIM is a completely standards-based Java EE web application with a zero client footprint. It is event-driven, providing a great deal of flexibility and control over information management processes. Access rights and ownership of data is role-based and can leverage existing LDAP systems. Business processes and associated rules are stored as XML and can be customized via the application GUI or directly by a developer. Communication with the outside world is done over standard protocols such as FTP, JDBC, web services, XML over JMS, or AS/2. Monitoring can be done by a standard JMX console. No specialized or proprietary

Skill sets are required for the implementation or management of TIBCO CIM.