Enterprise Application Integration

1
Enterprise Application Integration
2
Objectives

• ODS Overview
• EAI concepts
• EAI functional components

3
Enterprise Application Integration

• Enterprise Application Integration (EAI) is one
  of the hot-button issues in Information
  Technology in 2000. Information Week's research
  survey of 300 technology managers showed nearly
  75 of respondents said EAI is a planned project
  for their IT departments in the coming year.
• According to a study by Bank Boston, the market
  for EAI is expected to be 50 Billion USD in 2001

4
Enterprise Application Integration

• EAI began at Goldman Sachs in New York nearly 10
  years ago, where they funded the Teknekron
  Information Bus (TIB) to pump stock market quotes
  into different systems.
• The programmers who wrote Teknekron then left and
  founded TIBCO.
• Many of those same developers are now with
  Vitria.
•                          

5
ODS

• The operational data store (ODS) is a
  subject-oriented, integrated, current and
  volatile collection of data used to support the
  tactical decision-making process for the
  enterprise.
• Comparison ODS vs Data Warehouse

6
ODS

• The ODS retains much less data than the data
  warehouse
• ODS data ages (loses its value) very quickly
  typically in a few days

7
ODS

• The focus of the ODS is current transactions.
• ODS data is homogenous as compared with the data
  warehouse

8
ODS

• ODS data is volitile it is constantly updated as
  transactions occur in operational systems

9
Why build an ODS

• Rationale is similar to that of Data Warehouse
• Off load reporting and analysis functions from
  operational systems.
• Allow analysis of integrated data from across the
  enterprise
• Multiple web servers
• Multiple order systems
• Provides conformed data to users
• Provides proper analysis tools to users
• Facilitates sharing of value-added information
  between business functions (ordering and
  marketing)

10
Enabling the ODS

• The ODS relies on the receipt of data from
  operational sources in near-real time.
• A class of information system, Enterprise
  Application Integration (EAI) can help.

11
Enterprise Application Integration (EAI)

• Provides the means to share data between
  different applications without writing custom
  interfaces.
• Example
• Consider a company that wants to do build an
  eCommerce portal.
• Has a number of legacy applications (or even an
  ERP solution in place)
• Need build the web-based eCommerce infrastructure
  and link with systems that do functions like
  process orders, manage inventory, ship products
• The company decides to use a major courier
  service for delivery of products ordered from the
  web site.
• The company wants an ODS for analysis of
  transactions taking place on the site

12
Example
eCom Site
Order System
Inventory
ODS
Shipping
Courier
One solution custom Interfaces Issues?
13
Example
eCom Site
Order System
EAI
ODS
Inventory
Shipping
Courier
Alternative employ EAI solution
14
Advantages of EAI

• Cost effectiveness
• Time to market
• Scalability
• Ability to deal with complex environments

15
EAI Components

• Business Rule Component to allow the
  applications to understand your business
  processes
• Business Logic Modules (i.e. supply planning,
  sales order processing. Methods for business
  process management.)
• Transformation tools (to define how to map data
  from one system to another)
•                              

16
EAI Components

• Data Acquisition Component to allow access to
  the data Data Source and Target Interfaces (i.e.
  Siebel, SAP, PeopleSoft, ODBC, Oracle, CICS, IMS)
  - note that the data acquisition component is
  crucial to EAI success. Most vendors refer to
  these interfaces as "adapters"
• Adapters understand the data structures
  associated with applications and the means by
  which to access the data. (SAP/ABAP)

17
EAI Components

• System Development Component to allow
  programmers to design and test custom
  requirements - Design tools (for business process
  design, debugging, and testing)

18
EAI - Components

• System Control Component Should have the
  following features
• Management tools (for application-specific
  monitoring)
• Directory tools (for locating other applications
  on different platforms), particularly support for
  the Lightweight Directory Access Protocol (LDAP)
• Commitment control management mechanisms (for
  control of business-level logical units of work)
• Strong support for metadata management

19
EAI - Components

• Message Brokers (to control transactions, control
  security, and perform event notification. The
  product should also include the capability to
  "bridge" messages between different messaging
  systems
• Scalability for high-volume transaction
  throughput. It is almost impossible to know at
  implementation time what the data volumes will be
  in the future therefore, EAI must be scalable.
• Support for varying levels of fault tolerance,
  load balancing, and failover for mission-critical
  systems
• Workflow enablement is a key requirement to
  reduce latency between distributed processes.

20
Example

• Save for Later (Computer Makers Web Store)
• Potential client goes to the web site and
  registers
• Client then browses through products, placing in
  shopping cart.
• If client decides not to buy at a point in time,
  can Save For Later keep the order for a
  period and come back later to purchase.
• Once a SFL transaction takes place, the computer
  maker routes to an ODS, does some analysis.

21
Example

• SFL data may be used in a variety of ways
• Customer may be sent an email
• Transactions may be routed to call center for
  follow up (with customer registration data
  email, phone, etc) offer discounts, product
  information, etc. try to convince the customer
  to purchase.

22
Example

• How it works
• Save for later transaction on the web causes
  INSERT or UPDATE in certain DB tables
• Trigger fires after INSERT or UPDATE and routes
  data to EAI queue.
• EAI software reads the data in the queue does
  some simple transformations (mapping of data in
  this case) and forwards to ODS based on a set of
  rules (some transactions are ignored, for
  example)