Rapid Integration Tools For Rapid Application Development A Case Study on Legacy

Integration Amit Midha Ravindra Singh Lakshmi Pratha Hari Patrick R. Place, Advisor

December 2004

TECHNICAL REPORT CMU/SEI-2004-TR-023 ESC-TR-2004-023

Pittsburgh, PA 15213-3890

Rapid Integration Tools for Rapid Application Development

A Case Study on Legacy Integration

CMU/SEI-2004-TR-023 ESC-TR-2004-023 Amit Midha Ravindra Singh Lakshmi Pratha Hari Patrick R. Place, Advisor

December 2004

Integration of Software-Intensive Systems Initiative

Unlimited distribution subject to the copyright.

This report was prepared for the

SEI Joint Program Office ESC/XPK 5 Eglin Street Hanscom AFB, MA 01731-2100

The ideas and findings in this report should not be construed as an official DoD position. It is published in the interest of scientific and technical information exchange.

FOR THE COMMANDER

Christos Scondras Chief of Programs, XPK

This work is sponsored by the U.S. Department of Defense. The Software Engineering Institute is a federally funded research and development center sponsored by the U.S. Department of Defense.

Copyright 2005 Carnegie Mellon University.

NO WARRANTY

THIS CARNEGIE MELLON UNIVERSITY AND SOFTWARE ENGINEERING INSTITUTE MATERIAL IS FURNISHED ON AN "AS-IS" BASIS. CARNEGIE MELLON UNIVERSITY MAKES NO WARRANTIES OF ANY KIND, EITHER EXPRESSED OR IMPLIED, AS TO ANY MATTER INCLUDING, BUT NOT LIMITED TO, WARRANTY OF FITNESS FOR PURPOSE OR MERCHANTABILITY, EXCLUSIVITY, OR RESULTS OBTAINED FROM USE OF THE MATERIAL. CARNEGIE MELLON UNIVERSITY DOES NOT MAKE ANY WARRANTY OF ANY KIND WITH RESPECT TO FREEDOM FROM PATENT, TRADEMARK, OR COPYRIGHT INFRINGEMENT.

Use of any trademarks in this report is not intended in any way to infringe on the rights of the trademark holder.

Internal use. Permission to reproduce this document and to prepare derivative works from this document for internal use is granted, provided the copyright and "No Warranty" statements are included with all reproductions and derivative works.

External use. Requests for permission to reproduce this document or prepare derivative works of this document for external and commercial use should be addressed to the SEI Licensing Agent.

This work was created in the performance of Federal Government Contract Number F19628-00-C-0003 with Carnegie Mellon University for the operation of the Software Engineering Institute, a federally funded research and development center. The Government of the United States has a royalty-free government-purpose license to use, duplicate, or disclose the work, in whole or in part and in any manner, and to have or permit others to do so, for government purposes pursuant to the copyright license under the clause at 252.227-7013.

For information about purchasing paper copies of SEI reports, please visit the publications portion of our Web site (http://www.sei.cmu.edu/publications/pubweb.html).

CMU/SEI-2004-TR-023 i

Table of Contents

Background and Acknowledgements...................................................................vii

Abstract .................................................................................................................. ix

1 Introduction..................................................................................................... 1 1.1 Purpose and Objective ............................................................................. 1 1.2 Project Requirements ............................................................................... 3 1.3 Project Plan and Tracked Report .............................................................. 4 1.4 Structure of the Document ........................................................................ 4

2 Identification and Classification of Tools...................................................... 7 2.1 List of Tools............................................................................................... 7 2.2 Tool Selection Criteria............................................................................... 7 2.3 Classification Parameters ......................................................................... 8 2.4 Tool Evaluation ........................................................................................15

3 Evaluation Using a Model Problem ..............................................................17 3.1 Purpose ...................................................................................................17 3.2 Model Problem Selection.........................................................................17

3.2.1 Model Problems ...........................................................................17 3.2.2 Problem Selection ........................................................................18

3.3 Model Problem Description......................................................................20 3.4 Tool Evaluation using Model Problem ......................................................22

4 Conclusions ...................................................................................................29 4.1 Lessons Learned .....................................................................................29 4.2 Future Directions of the Research ...........................................................30 4.3 Remarks ..................................................................................................30

Appendix A Tool Studies and Analysis..............................................................31

Appendix B Tool Evaluation Reports .................................................................45

Appendix C Model Problem and Analysis .........................................................59

ii CMU/SEI-2004-TR-023

Appendix D Commercial Off-the-Shelf Components ........................................69

Appendix E Trading Bond System COCOTS Estimation Details .....................75

Appendix F Project Details.................................................................................81

Glossary of Technical Terms.................................................................................85

References/Bibliography.......................................................................................91

CMU/SEI-2004-TR-023 iii

List of Figures

Figure 1: Evaluation Process for the Rapid Integration Tools................................. 2

Figure 2: Evaluation Framework............................................................................ 3

Figure 3: Graph Showing Characteristics of the Three Tools Selected .................16

Figure 4: High-Level Context Diagram of Trading Bond System...........................20

Figure 5: Structure of the Model Problem .............................................................22

Figure 6: Model Solution—High-Level Context Diagram.......................................24

Figure 7: Graph that Explains the Estimated vs. Actual Effort and Cost................27

Figure 8: Legacy Market Data Subsystem............................................................60

Figure 9: Legacy Contribution Subsystem ............................................................61

Figure 10: Logical View of the System....................................................................65

Figure 11: Feeder Component Specification...........................................................70

iv CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 v

List of Tables

Table 1: Classification Parameters - Technical and Non-Technical ....................... 8

Table 2: Weights Assigned to Parameters Based on Rules of Thumb .................10

Table 3: Prioritized List of Quality Attributes ........................................................22

Table 4: Top Three Risk List................................................................................25

Table 5: Variance Calculator ..............................................................................26

Table 6: Posteriori Evaluation Criteria Satisfied by the Tools ...............................27

Table 7: Tools Observations Conforming to Non-Functional Requirements.........28

Table 8: Milestones and Schedule of the Project.................................................82

vi CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 vii

Background and Acknowledgements

The project served as an educational elective that is a requirement for the PDC [Professional Devel-opment Center] Scholars at Carnegie Mellon University-West Coast Campus for their MSIT-SE Curriculum.

We would like to thank Patrick Place, Member of the Technical Staff at the Software Engineering In-stitute, Pittsburgh, for mentoring the research and providing guidance and direction.

We would also like to thank those listed below for their approval and support.

• Dr. Lynn Robert Carter, Principal Fellow, West Coast Campus, Carnegie Mellon University

• Patricia Oberndorf, Director, Dynamic Systems Program SEI

• Dennis Smith, Lead, Integration of Software Intensive Systems Initiative

• Dr. Scott Lewis, Senior Faculty, West Coast Campus

Industry Experts:

• Gerry Miller, Chief Technology Officer, Microsoft Corporation, U.S. Central Region

• Eric Newcomer, Chief Technology Officer, IONA Technologies

• Richard. W. Metz, Director, Process, Architecture and Tools, Boeing Computer Services

• Jim Farrell, Product Management, IBM Rational Software

viii CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 ix

Abstract

This report investigates the rapid integration tools available in the current market. These tools aid in the rapid integration of software systems and components. The research centers on a model problem that requires such a tool to address legacy integration challenges. The report presents a generic evaluation framework for identifying and evaluating rapid integration tools and an evaluation of three identified tools. This evaluation engaged selected evaluation criteria based on the demands of the model problem. A process reference is also included; this forms the guidelines for identification and evaluation of the tools with respect to other model problems.

x CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 1

1 Introduction

1.1 Purpose and Objective This project involves the analysis of rapid integration tools available in the market, which aid in rapid integration of software systems/components. The project is centered on a model problem that requires such a tool to address legacy integration challenges. The main outcome of this research includes

• a generic evaluation framework for identifying and evaluating rapid integration tools. The evaluation criteria are geared towards the model problem that belongs to a class of model prob-lems having integration/interoperability as the key concern.

• an evaluation of three identified tools with respect to the evaluation criteria and the model prob-lem which forms the framework for evaluation of tools.

• a process reference to the Integration of Components Certificate at Carnegie Mellon West, which forms the guidelines for the identification and evaluation of the tools with respect to other model problems.

2 CMU/SEI-2004-TR-023

Figure 1: Evaluation Process for the Rapid Integration Tools

The above diagram symbolizes the process followed for determining the evaluation framework. The team identified the model problem and the list of tools, quantified requirements from the model prob-lem description, came up with a tool evaluation report and finally came up with an evaluation frame-work. The figure below illustrates the evaluation framework defined for the tools that have been se-lected to satisfy the requirements specified as critical by the model problem. In both the preceding and following diagrams technical factors are those directly related to the model problem and are de-rived from both functional and non-functional parameters. The non-technical parameters are softer, but no less important, factors such as the quality of vendor support or market share of the tool.

Requirements Gathered

Identify List of Rapid

Integration Tools

Identify Critical Requirements

(Integration Perspective)

Tool Evaluation Based on Technical and Non-Technical

Parameters

Evaluation Framework

Model Solution

Selected Tools

Model Problem Requirements Vs Tool Features

Selected Requirements

Identify Model

Problem

Featured Gathered

CMU/SEI-2004-TR-023 3

Figure 2: Evaluation Framework

1.2 Project Requirements As demand for new functionality grows and new systems to fulfill it go into operation, the need to integrate new systems with existing systems has increased. The development of resulting extended systems is frequently based on the integration of existing components, leading to demand for new integration tools. Modern integration tools all promise the ability to integrate components more quickly and cheaply than traditional technologies.

The project described here is aimed at surveying the field of rapid integration tools with a view to informing the reader on how to select among the choices. The task was divided into the following steps.

Survey and classify the tools.

The first step was to identify the tools that claim to provide rapid integration. Since these tools were known to provide a wide range of services, the identification also required the development of a clas-sification scheme for characterizing the various types of tools.

Deliverables: 1) list of rapid integration tools 2) classification scheme 3) classified list of tools

Technical Pa-rameters

Rapid Integration Tools

Tool Evalua-tion Results

Non Technical Parameters

Identified critical requirements for Model Solution

Weights Assigned based on Risk/Misfit

Rules of thumb applied on se-lection of pa-rameters

Model Problem Requirements

4 CMU/SEI-2004-TR-023

Evaluate the tools using a model problem

We expected that one or more of the classifications would contain a number of interesting rapid inte-gration tools. We chose a problem typical of the type of integration the tools were designed for and applied a number of tools to that problem.

Deliverables: 1) preliminary evaluation scheme 2) model problem definition 3) reports detailing evaluation of tools’ applicability to the model problem

Develop and document general evaluation criteria

Following the evaluation, the final step was to refine the evaluation criteria and document the refined versions. The purpose was the creation of an instrument that would assist a developer in choosing the “right” rapid integration tool.

Deliverable: documented evaluation criteria for rapid integration tools. Depending on time, steps 2 and 3 may be repeated within another classification.

1.3 Project Plan and Tracked Report We followed a simple phased approach for executing the project with each phase divided into tasks and related deliverables. Each deliverable is considered as a milestone and is derived from the initial list provided by the SEI. Since the project is exploratory, it does not follow any standard software development life cycle, but we followed software engineering principles from the start. We used the work breakdown structure (WBS) and effort available from the elective to estimate a completion date based on a given start date. The project ran over schedule perhaps indicating the problem of using available effort as an artificial constraint on work to be performed.

1.4 Structure of the Document This report is organized into three major chapters.

Chapter 1: Introduction to the Technical Report presents the purpose and objective of the project, the project description, background, requirements, project plan and tracked report and the structure of the document.

Chapter 2: Identification and Classification of the Tools describes the list of tools identified as the rapid integration tools and the evaluation framework applied to them for selection to work with the model problem. The classification parameters that support the evaluation framework are the technical and non-technical parameters.

Chapter 3: Model Problem and Tool Implementation explains the model problem selection and identification of critical requirements as well as application of the tools and their assessments.

CMU/SEI-2004-TR-023 5

Chapter 4: Conclusions documents the lessons learned arising from the use of the specific tools. Additionally, questions for future research are listed as are some concluding comments on the devel-opment of the evaluation framework, including factors to consider before and after applying the evaluation framework.

Appendices feature detailed descriptions of the tools evaluation, model problem, and other estima-tions.

6 CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 7

2 Identification and Classification of Tools

2.1 List of Tools The first step in our approach is to identify the tools used for rapid integration. We discovered little difference between integration tools and rapid integration tools.1 We identified 11 rapid integration tools designed for the rapid integration of applications from existing components.

1. Pervasive Data Junction

2. RoughWave's LEIF

3. IBM Rational Rapid Developer

4. Microsoft SQL Server

5. Host Integration Server

6. Microsoft BizTalk Server

7. IBM WebSphere Business Integration

8. Artix Relay, Encompass and Mainframe

9. PiiE Smart Client and Fusion Server

10. InterSystem Ensemble

11. Jboss

For the above-listed tools, we collected information about their vendors’ name and features. See Appendix A: Tools Studies and Analysis for more information.

2.2 Tool Selection Criteria The identified tools were filtered based on the model problem that will be described in Section 3. Since 11 tools seemed too many for starting the evaluation process, a short list was created based on the following criteria.

1. Tool should be capable of solving a wide range of Enterprise Application Integration problems, especially the Legacy Integration problem.

2. Tool is able to provide communication between Java and C++ Components.

3. Tool has solid success stories associated with it.

1 Rapid is a concept that depends on the user’s context. In some contexts, six months may be considered

rapid and in others, six hours could be too long. The tools themselves are, essentially, the same and a better question is whether the integration tools speed the integration sufficiently to both produce timely applica-tions and cost less than not using the tools.

8 CMU/SEI-2004-TR-023

4. Tool has been in market for at least two to three years.

5. An evaluation version of the tool is available and the evaluation period is sufficient to evaluate the tool.

Through application of these selection criteria the above list of 11 tools was short-listed to 3. The three tools were

1. IBM Websphere

2. IBM Rapid Developer

3. LEIF ( Light-weight Enterprise Integration Framework)

2.3 Classification Parameters The Classification Parameters used to evaluate tools can be technical or non-technical in nature. The functional and non-functional requirements of the model problem form the technical parameters. Powell and colleagues observed] that apart from these technical parameters, some non-technical pa-rameters arise from other business-oriented issues, such as cost and vendor, which play an important role in the selection of a tool for rapid application development [Powell 97].

We identified 16 parameters (5 non-technical parameters and 11 technical parameters) for classifying rapid integration tools. The table below gives a brief description of these parameters.

Table 1: Classification Parameters - Technical and Non-Technical

# Parameter Description Non-Technical Parameters

1 Business market price of the tool

return on investment (ROI) of the tool (based on cost of the tool compared to the estimated cost of manually integrating the components)

foreseen risk in using the tools (lifespan of the tool, ease of use, change frequency and so on)

2 Evaluation-Specific project life cycle in which the tool can be used (software con-figuration, project planning, oversight and tracking and so on)

comparative report of other tools in similar domain

3 External References visibility and popularity of the tool in the market

4 Vendor Support quality and cost of the vendor support

access to architecture and design aspects of the tool

CMU/SEI-2004-TR-023 9

5 Tool-Specific

integration with other tools and platform it can support

solution space the tool belongs to with respect to the problem

(domain specific)

reliability of the tool and the vendor maturity level based on

industry standards

skill set required to operate the tool

tailorability of the tool

extent to which data generated by the tool (performance logs

and so on) is configurable.

number of well-defined components that can be used

separately

performance of the tool

interactivity of the tool

sufficiency of documentation (user manual, installation guide

and so on) bundled with the tool

degree to which data generated by the tool can be used by

other tools

Technical Parameters

6 Security support offered by the tool for developing secure or safety

critical systems

7 Correctness capability of the tool for producing accurate results

8 Availability and

Robustness

capability of the tool for surviving system failure

9 Ease of Use - Usability degree of learning curve associated with the tool

10 Downward and portability of applications developed using one version of the

10 CMU/SEI-2004-TR-023

Upward Compatibility tool to higher and/or lower versions of the same tool

11 Flexibility capability of tool for operating in different operating system

environments

12 Product Performance response time of the tool

13 Tailorability customizability of tool for meeting user-specific requirements

(user interface, enabling/disabling of features, enhancing the

tool by adding plug-ins, and so on)

14 Service Implementa-

tion Coverage

technical support/licensing cost associated with the tool

15 Interoperability capability of tool to interoperate with other systems

16 Testability ability to test the functionality of the tool

Each rapid integration tool is analyzed based on the classification parameters above; it is rated on a scale of 0 to 10, depending on how well it satisfies the parameters. The detailed evaluation of the tools is found in Appendix A: Tool Studies and Analysis.

Assigning values to parameters while analyzing any tool may be tricky. Different individuals may come up with different analysis results. In order to avoid this, we defined some rules of thumb, shown in Table 2 below. These rules are so generic that they can be used to analyze any rapid integration tool.

Table 2: Weights Assigned to Parameters Based on Rules of Thumb

Weights

Parameters 0 1 to 3 4 to 7 8 to 10

CMU/SEI-2004-TR-023 11

Business Cost is very high and has special installation re-quirements (e.g., specific opera- ting system, run- time libraries).

Cost of the tool doesn't support the ROI; there are fre-quent changes to the tool and the cost of learning the tool is high.

Cost of the tool sup-ports the ROI; there are frequent changes to the tool and the cost of learning the tool is high.

Cost of the tool supports the ROI; there are two or three releases of the tool a year and the cost of learning the tool is low (e.g., because of exten-sive Graphical User Interface).

Evaluation- Specific

Tool is single-user and supports no integration with organization's software devel-opment life cycle and other tools.

Tool assists in the collaborative devel-opment but cannot be integrated with the organization's software develop-ment life cycle.

Tool supports collabo-rative development by team, has its own con-figuration manage-ment and project management utility but cannot be inte-grated with other tools.

Tool supports col-laborative develop-ment by team, sup-ports configuration and project man-agement and can be integrated with other tools to ex-pand its current ca-pabilities.

External References

Tool has recently launched in the market.

Tool has received average response from the user, has been in market for one to two years, and a similar tool by leading vendors (e.g., Microsoft, IBM) is available in the market.

Tool has been used by several large organiza-tions, has very few competitors, and has several success stories associated with its use.

Tool has been in market for four or more years, owned by software market leaders like IBM or Microsoft, used by large organizations, and has many suc-cess stories associ-ated its use.

Vendor Support

Tool has no cus-tomer support.

Tool has limited customer support through mail and telephone conversa-tions only.

Tool has good cus-tomer support through online discussion fo-rum, mail and tele-phone conversations. There is immediate response to queries posted to Customer

Tool has effective customer support through online dis-cussion forum, email, and on-site consultation. Re-sponse is immediate to queries posted to

12 CMU/SEI-2004-TR-023

Support Center. Customer Support Center.

Tool-Specific Tool doesn't sup-port integration with other tools.

Tool supports inte-gration with two or three other tools and has complex inte-gration process.

Tool supports integra-tion with software configuration man-agement tools, testing tools, application servers, and so on, and integration proc-ess is moderate and requires manual set-tings.

Tool supports inte-gration with soft-ware configuration management tools, testing tools, appli-cation servers and so on, and integra-tion process is eas-ily performed via wizards. Tool sup-ports custom devel-opment to enhance its features and us-ability.

Security Tool doesn't pro-vide any features to aid in the im-plementation of security mecha-nism (encryption, authentication, authorization etc.)

Tool supports few standard security mechanisms like encryption and au-thentication.

Tool supports most security mechanisms currently used in the market and but doesn't support any custom development of secu-rity mechanisms.

Tool supports most security mecha-nisms currently used in the market and also supports custom develop-ment of security mechanisms.

Correctness Tool has no utility for testing the ap-plication devel-oped by it.

Tool supports lim-ited testing for the application devel-oped.

Tool supports stan-dard testing of the application developed through testing utili-ties bundled with the tool.

Tool performs vali-dation at every step while developing the application. Also supports inte-gration of other test-ing tools (e.g., third party application servers) to verify the correctness of the application cre-ated.

CMU/SEI-2004-TR-023 13

Availability and Robustness

Tool doesn't save the data should system failure occur.

Tool backs-up the application data, which it uses to re-cover from system failure.

Tool backs up the ap-plication data and provides automatic recovery from any type of system failure (sudden shutdown of the desktop, sudden crashing of desktop etc.).

Tool takes backs up of the application data, provides automatic recovery from any kind of system failure (sud-den shutdown of the desktop, sudden crashing of desktop etc.) It also supports restoration points so that user can switch between restoration points.

Usability Tool has non-GUI interface and no feature to auto-mate the execu-tion of tasks or operations.

Tool has GUI Inter-face, but requires a lot of navigation across the screen to perform any opera-tion.

Tool has GUI Inter-face with minimum overhead of naviga-tion while performing any task. Also it pro-vides quick links to commonly used op-erations.

Tool has effective GUI Interface which not only eases in performing tasks but also re-duces the learning curve associated in performing any task. Also tool sup-ports has wizards to guide operations step by step and single-click execu-tion of any opera-tion.

Upward and Downward Compatibility

Application cre-ated by the tool is not supported by earlier or newer versions of the same tool.

Application created by the tool can only be exported to new version under a few circumstances.

Application created by the tool can be ex-ported to new version but requires manual changes in the con-figurations.

Application created by the tool can be exported to new versions. All the necessary changes are automatically handled by the tool itself.

14 CMU/SEI-2004-TR-023

Flexibility Tool has a specific Operating System requirement.

Tool is available for several Operating System environ-ments and does not support portability of applications be-tween platforms.

Tool is available for several Operating System environments and supports portabil-ity of applications between platforms.

Tool is available for several Operating System environ-ments, supports portability of appli-cation between plat-forms and has inter-faces for communication be-tween instances running on different platforms.

Product Performance

Time taken by tool to perform any operation is more than six minutes and the system hangs up while performing any operation.

Time taken by tool to perform any op-eration is four to six minutes and 80% of the time tool per-forms its operation successfully.

Time taken by tool to perform any operation is two to four minutes and 90% of the time tool performs its op-eration successfully.

Time taken by tool to perform any op-eration is between two to four minutes and 100% of the time tool performs its operation suc-cessfully.

Tailorability Tool doesn't allow user to configure / enhance its features.

Tool allows user to configure / enhance its features by in-stalling plug-ins or add-ons available from the tool ven-dor only.

Tool allows user to configure / enhance its features by install-ing suitable plug-ins or add-ons available from any vendor.

Tool allows user to configure / enhance its features by in-stalling suitable plug-ins or add-ons available from any vendor or by pro-gramming the tool itself.

CMU/SEI-2004-TR-023 15

Service Implementa-tion Coverage

Tool has stringent licensing policy and does not pro-mote evaluation copies to experi-ment with the tool. Also it has high licensing cost and purchasing a new license is almost equal to the cost of the tool itself.

Tool promotes evaluation copy but the period is not sufficient enough to evaluate the tool. Also the licensing cost is very high.

Tool promotes evalua-tion copy with ade-quate customer sup-port for any issues that arise during the evaluation period but the evaluation period is not sufficient enough to evaluate the tool. The licensing cost is nominal.

Tool promotes evaluation copy with adequate cus-tomer support for any issues that arise during the evalua-tion period and also the evaluation pe-riod is sufficient enough to evaluate the tool fully. The licensing cost is nominal.

Interopera-bility

Results produced by the tool cannot be exported to other formats (Word document, html, jpeg, etc.) and it doesn't have any interface to communicate with other tools.

Results produced by the tool can be ex-ported to other for-mats (Word docu-ment, html, jpeg, etc.) but it does not support inter-tool communication.

Results produced by the tools can be ex-ported to other for-mats (Word docu-ment, html, jpeg, etc.) and it supports inter-tool communication.

Results produced by the tools can be ex-ported to other for-mats (Word docu-ment, html, jpeg, etc.) and it supports inter-tool communi-cation. Also tool can produce results that can be ported to any platform.

Testability of Output

Tool doesn't vali-date the output produced.

Tool does minimal validation and the accuracy of the re-sult is about 70%.

Tool does validation of the results and the accuracy of the result is about 80%.

Tool does validation throughout and the accuracy of the re-sult is 100%.

2.4 Tool Evaluation The three short-listed tools were evaluated for these technical and non-technical parameters. See Appendix A: Tool Studies and Analysis for the Tool Evaluation Report.

The following graph shows the summary of the parameter values for each tool.

16 CMU/SEI-2004-TR-023

Tools v/s Parameter

0123456789

10

Busine

ss

Evaluation Spe

cific

Para

meter

s

Exter

nal R

efer

ence

s

Vendo

r Sup

port

Tools

Spe

cific

Secur

ity

Corre

ctnes

s

Availa

blity

& Rob

ustnes

s

Usability

Upwar

d &

Downw

ard

Compa

tibilit

y

Flex

ibilit

y

Produ

ct Pe

rform

ance

Under

stand

ibility

Tailo

ribilit

y

Servic

e Im

plemen

tatio

n Cov

erag

e

Inte

rope

rabil

ity

Testab

ility

Parameter

Wei

ghts IBM Websphere

LEIF

IBM Rational Rapid Developer

Figure 3: Graph Showing Characteristics of the Three Tools Selected

The above graph shows the comparative analysis of three rapid integration tools—IBM Websphere, IBM Rapid Developer, and LEIF—with respect to technical and non-technical parameters. The X-axis of the graph lists the parameters (technical and non-technical) and the Y-axis represents the value assigned to these parameters from 0 to 10. Such graphs can be used to prioritize the list of identified rapid integration tools. A similar graph including model problem requirements in terms of parameters can help us to identify which tool is the best fit for the model problem.

CMU/SEI-2004-TR-023 17

3 Evaluation Using a Model Problem

3.1 Purpose This section describes the evaluation of the tools using a selected model problem. The model problem selection criteria, the description of the model problem, and the evaluation of the tools are high-lighted. In describing the model problem selection criteria we also explain the sequence of steps we followed in making our selection. We briefly describe the model problem with respect to the non-functional requirements and the problem statement. Finally, we explain the tool evaluation using the model problem; here we’ve applied the structure of the selected model problem as described by Kurt Wallnau in Building Systems from Commercial Components [Wallnau 02]. The assessment results obtained from this evaluation show the extent to which the tools satisfied the post evaluation criteria and the problem’s non-functional requirements.

3.2 Model Problem Selection

3.2.1 Model Problems

We found three potential model problems (descriptions follow); each problem is appropriate to a par-ticular type of integration.

Web Service Enablement

A company uses the enterprise integration technology as well as XML technology to make customer account information accessible via a Simple Object Access Protocol (SOAP)-based Web services in-terface.

The implementation of this solution requires retrieving and combining information from two source applications. The first application is a custom CORBA system that provides historical customer sup-port information. The second application is Siebel, which provides customer purchase information.

Legacy Integration

Bond traders working online must send prices for a large number of bonds to several different trading venues, each with its own user interface; this disrupts the workflow of their bond trading desk.

The system solution should minimize the minutiae of pricing all traders’ bonds and provide an ad-vanced analytic functionality, specific to the bond market, in a single encapsulated user interface. This system would utilize legacy components on the server side.

18 CMU/SEI-2004-TR-023

Business Application Integration

Three companies use different business processes involving different sets of assumptions. Middle-ware must be utilized as the integration point for communicating among the business processes.

3.2.2 Problem Selection

The following five steps were followed

Step 1: Identify the problem that must be solved through integration.

In this case we identified the following types of integration based on our prior knowledge.

1. Middleware Integration

2. Service Oriented Integration

3. Web Service Integration

4. Legacy Integration

5. Enterprise Application Integration

Step 2: Based on research on the various application integration types we chose three that would pro-vide the best opportunity for using the tools that we have selected.

1. Legacy Integration

2. Business Application Integration

3. Web Service Enablement

Step 3: Analysis and study on the three model problems were made based on answering the following questions.

1. Which problem gives a way to integrate two different technologies?

Of the three problems presented, we found that the Trading Bond System required a solution that would integrate components built using two different technologies (in this case programming lan-guages), as evidenced by its case study report:

Traders needed a very responsive application on both Windows NT and Solaris workstations. Therefore, we decided to implement the client application as a Java thick client because of its platform independence and its ability to quickly respond to user input and market data. On the server side, we are inheriting legacy C++ components that our system will utilize [Simon 03].

The Trading Bond system meets the criterion of providing a way to integrate two technologies. The solution requires integration of a Java based component and C++ based components, which can be done by building a pair of Java gateways to communicate with the C++ server-side components [Simon 03]. Details of the Trading Bond System case are available at http://www.eaipatterns.com/BondTradingCaseStudy.html.

CMU/SEI-2004-TR-023 19

2. Which problem is faced in the real world often?

A demand exists in government and industry for existing systems to be updated or integrated with the current technology. Many applications require technology independence and interoperability with various applications that were developed using different technologies. Moreover, in the current indus-trial scene there is a drive toward production of tools for the integration of Java-based components and C++ based components. The Trading Bond System is therefore quite typical of real-world scenar-ios.

3. Which problem is very specific and solvable within the specified time constraint?

We found that legacy integration for the Trading Bond System best met these criteria. We were able to download two specific components provided by Dukascopy, an online trading application, which mapped with the Trading Bond System scenario. This bode well for solving the problem on time.

We did not find the same with the customer account information problem to be solved through Web Service enablement: Here the application was generic and we were not able to find the specific attrib-utes to be addressed or specific requirements to be fulfilled. We realized that it might take consider-able time to establish which tools were needed. This presented a problem, given time constraints and resource availability. We did not have enough time to create a simulation of the CORBA System and the Siebel system.

The problem that might have been solved through Business Integration did not meet this criterion. This involved three companies who required communication among their different business proc-esses. The business processes to be integrated were not well defined or specific. The time required for creating simulated processes and then integrating them did not meet our constraints.

4. Which of the other classes of integration does this model problem address?

The chosen problem could also be used to assess service-oriented integration insofar as it is reason-able to treat the problem components as services. Additionally, the middleware, application and Web-based integration classes could be addressed by the Trading Bond system.

Step 4: Identify the model problem that fit into the evaluation framework based on the above identi-fied questions.

Step 5: Having identified the model problem to be solved, we now present details regarding the Trad-ing Bond system relevant to further application development using the integration tools.

20 CMU/SEI-2004-TR-023

3.3 Model Problem Description The Trading Bond system best met the problem criteria and highlights the necessity of integration, for communication purposes, between various components with various user interfaces and communica-tion protocols. The following table shows an analysis of the original problem statement into a prob-lem description describing the various actors and also constraints on the solution.

The problem of bond traders to send prices for a large number of bonds to several different trading venues, each with its own user interface

affects bond traders

the impact of which is to disrupt the streamline of the workflow of their bond trading desk

a successful solution would be a bond pricing system to minimize the minutiae of pricing all of their bonds combined with advanced analytic functionality specific to the bond market in a single encapsulated user interface.

Figure 4: High-Level Context Diagram of Trading Bond System

First, market data comes into the system. Market data is data regarding the price and other properties of the bond representing what people are willing to buy and sell the bond for on the free market. The market data is immediately sent to the analytics engine that alters the data. Analytics refers to mathe-matical functions for financial applications that alter the prices and other attributes of bonds. These are generic functions that use input variables to tailor the results of the function to a particular bond. The client application that will run on each trader desktop will configure the analytics engine on a per trader basis, controlling the specifics of the analytics for each bond the trader is pricing. Once the

CMU/SEI-2004-TR-023 21

analytics are applied to the market data, the modified data is sent out to various trading venues where traders from other firms can buy or sell the bonds.

The following are some of the non-functional requirements that the system should address, in order of priority. The scope of this model problem extends to only the highest priority quality attributes se-lected by the team.

Integrability:

“On the server side, we are inheriting legacy C++ components that our system will utilize.”

The system should be integrable with the legacy C++ components which forms the Market data feed pricing subsystem and the thick client application which will be a Web-based thick Java client.

Performance:

“Traders need a very responsive application”

Two attributes of performance are essential to this responsiveness.

1. Scalability: Measured as the number of traders who will be accessing the system and the sys-tem’s capability for accommodating them.

2. Response Time: The system should be able to respond to the user without significant delay [here we say less than 5 sec, assuming that it is a Web-based application]

Portability:

“Traders need a very responsive application on both Windows NT and Solaris workstations.”

The application should be portable to any platform based on the demands of the trader’s needs.

The quality attribute that will be addressed in this execution of the model problem is highlighted in Table 3.

22 CMU/SEI-2004-TR-023

Table 3: Prioritized List of Quality Attributes

3.4 Tool Evaluation using Model Problem The following diagram shows the elements of the assessment. The trading bond problem is used as the model problem and the criteria coupled with the design question lead to the tool assessment, and COTS components forming the trading bond system.

Figure 5: Structure of the Model Problem

Design Question:

This is the initiating element of the model problem.

In this case study of the Trading Bond system, the design question is

Quality Attribute

Prioritization Rationale

Integrability 1 Use of the rapid integration tools to integrate legacy systems with Web-based application. In this case we are integrating the C++ legacy system with the Web-based Java Client.

Performance 2 We address the response time specific to this application as de-termined by the team for a responsive application.

Portability 3 Client application is developed in Java which automatically sup-ports platform independence.

Design Question [Hypothesis]

Priori Evaluation Criteria

Posteriori Evaluation Criteria

Minimum Relevant Constraints

Assessment Results

Trading Bond System

CMU/SEI-2004-TR-023 23

Is it possible to integrate the Java and legacy C++ components that are obtained off the shelf from the Dukascopy stock quote Web site, using the rapid integration Tools?

Hypothesis: A wrapper component [integration point] that provides the communication between the Java thick client and the legacy server side C++ component can be constructed using the rapid inte-gration Tools.

Priori Evaluation Criteria:

These are the criteria to be satisfied by the model solution. They were obtained by analyzing the ap-plication specifics given in the case study. They are centered on integration techniques and use of in-tegration tools. These evaluation criteria are formulated based on the hypothesis that we have ad-dressed related to the model problem. These criteria help in defining with the Standard’s compliance that the tools must meet in order to satisfy the requirements.

Criterion #1: A Java to C++ translator is required. Java thick Client talking with C++ Legacy Servers

Criterion #2: Messaging Bridge to support the communication between cross-language applications [C++ and Java]

Criterion #3: Single point of access is required to communicate with the gateways of the legacy serv-ers.

The criteria form the model problem requirements for the integration implementation using the tool. Thus according to Criteria #1, #2, and #3, the tool should be able to provide a communication mechanism, a messaging bridge and a single point access between the Java and C++ components. In this case the tools IBM WebSphere and the LEIF help in achieving these developments.

Minimum Relevant Constraints

The following constraints are based on what is feasible to provide in the model solution to address the above mentioned priori criteria:

1. This is a short-term project that involves rapid development; hence the use of rapid integration tools to create the Java to C++ translator, messaging bridge and single point access mechanism, which are the priori criteria of the model problem.

2. The business processes of the model problem are not a focal point, since they are addressed by the off-the-shelf components that are downloaded from the Dukascopy site.

3. The tool selection is restricted to the major functionalities provided by the tool with respect to the model problem’s priori criteria.

The development and deployment environment are the same; hence the performance of the model solution is constrained.

24 CMU/SEI-2004-TR-023

Model Solution: Trading Bond System

A simple solution that clarifies how the model solution was implemented is provided below:

Figure 6: Model Solution—High-Level Context Diagram

The Java Gateways are considered the Service Requestor and the Java Web Services are implemented using the IBM WebSphere. These Java Web Services are the Client side application required for communicating with the C++ Legacy servers, in this case the Market Data Feed Component obtained from the Customized Dukascopy Data Feed (CDDF) http://www.dukascopy.com/english /ddf_main/ . The C++ Legacy Servers are the service providers. The inner workings of the C++ Component and the Java Component were not considered; it was the integration between these com-ponents that was implemented using the Tools. The Discovery Agencies used were the UDDI Ser-vices, which were automatically set in the IBM WebSphere tool.

The System uses the simple publish-subscribe model for the implementation of the integration through discovery agencies and SOAP is the communication protocol that establishes the interaction between the two Web services.

Posteriori Evaluation Criteria

Criterion #1: Installation and development environment for the identified solution tools are in place.

Criterion #2: The off-the-shelf components architecture and design maps with the model problem requirements.

Criterion #3: The integration of the two COTS components is accomplished using the rapid integration tools

Service

Requestor 1

Service

Requestor n

Client

Service

Provider

Client Services

C++ Legacy Servers

Find

Java Gateways

Discovery

Agencies

Publish

Interact

CMU/SEI-2004-TR-023 25

The above criteria help in evaluating whether the tools are able to meet the requirements of the model problem and whether they are able to conceptualize the hypothesis that has been defined for this model problem.

Assessment Results

The assessment results are enumerated based on the following factors:

1. risks encountered and mitigated while using the contingency approach

2. the size, effort, and cost variance involved when using the rapid integration tools and when not using the rapid integration tools

3. product outcome explaining the steps that brought success and those that resulted in failure in the development using rapid integration tools

These assessments help in evaluating the tools as they apply to the model problem. In this case they are restricted with respect to the legacy integration of cross-language platforms.

Risks Encountered and Mitigated

The following table describes the major risks that we encountered and mitigated through contingency plans.

Table 4: Top Three Risk List

No. Risk Risk Management Strategy Status

Contingency

Impa

ct

Occ

urre

nce

Pro

babi

lity

Exp

osur

e

(Ran

k) Mitigation

Activities Trigger Activities

1 Tools iden-tified are not suitable for solving the model problem

90% .9 1 Collect the tools based on the model problem’s critical re-quirements

The tools are not able to produce a mechanism that solves the commu-nication be-tween Java Gateways and the C++ Legacy Server.

Determine which tools support the communication mechanism. In this case C++ Web Services are created using the Apache Axis C++. The integra-tion of the C++ Component with the Java Compo-nent is accom-plished via the LEIF.

Close

26 CMU/SEI-2004-TR-023

No. Risk Risk Management Strategy Status

Contingency

Impa

ct

Occ

urre

nce

Pro

babi

lity

Exp

osur

e

(Ran

k) Mitigation

Activities Trigger Activities

2 Learning curve

60% .8 2 Estimate the effort and exe-cute a short-term plan for learning only the required tools.

Understand-ing the proc-ess of using the tool for the specified model prob-lem.

Approach the technical support for the specific tool or the inter-active manual for the understanding of the tool.

Close

3 Installation and trou-bleshooting

80% .7 3 Test the devel-opment envi-ronment using Evaluation Software and samples.

Installation is problematic or the tool is unable to produce the required functionality.

Use separate test-ing machine for testing the instal-lation and run sample problems that are related to the model prob-lem requirements

Close

Size, Effort, Cost Variance

The following table explains the size, effort and cost variance. The size, effort and cost are estimated using the COCOTS calculator; this includes estimates of the glue code to be written and calculation of the respective effort and cost for writing the glue code. Using the actual size, effort, and cost re-corded while doing the development, the variance is calculated as shown below:

Table 5: Variance Calculator

Factor Estimated without using the rapid in-tegration Tools

Actual after using the rapid integration Tools

% Variance (Estimated-Actual)/Estimated *100)

Size (KLOC) 1.01 0 [Source Code auto generated]

100

Effort (Person Months) 17.63 8 54.6

Cost ( in $ excluding software costs)

123,403 55,996 54.6

CMU/SEI-2004-TR-023 27

Estimated Vs Actual

0

20

40

60

80

100

120

140

Size Person Month Cost

Factors

Estimated

Actual

Figure 7: Graph that Explains the Estimated vs. Actual Effort and Cost

Product outcome

The product developed using these tools should have adhered to the a posteriori evaluation criteria that we arrived at and also the non-functional requirements of the model problem.

Table 6: Posteriori Evaluation Criteria Satisfied by the Tools

Criterion #

Description Observation

1 Installation and development environments are in place for the identified solution tools.

Yes. All three tools satisfied this criterion.

2 The off-the-shelf components architecture and design satisfy the model problem re-quirements.

The components do not map exactly with respect to the implementation model as required by the model prob-lem. They satisfied the functional requirements.

3 The integration of the two COTS compo-nents is accomplished using the rapid inte-gration tools.

The tools, especially IBM Web-Sphere, LEIF, and Apache Axis C++, were used for creating and integrating the Web services of the COTS components.

The product outcome is also validated when the non-functional requirements are satisfied by the vari-ous tools.

28 CMU/SEI-2004-TR-023

Table 7: Tools Observations Conforming to Non-Functional Requirements

Requirement LEIF IBM Websphere IBM Rapid Devel-oper

Integrability Provides services for integration rather than integration itself and is limited re-garding C++ Tech-nology

Provides integration capability and is limited regarding Java Technology

Integration capabil-ity is lower and is limited regarding Java Technology

Performance (development time provided by the tool, not involving the prerequisites)

Simple interface with fewer inputs and quick response (2 minutes)

Requires knowledge about Web services and complex user interaction and is highly responsive (4 minutes)

Ease of use, and good user interface, and good response time (3 minutes)

Portability (based on the platform inde-pendence of the tool)

Portability is very possible (able to cre-ate services for vari-ous operating sys-tems).

Portability is not possible. Caters to only J2EE applica-tions / middleware applications

Portability is sup-ported to a limited extent.

CMU/SEI-2004-TR-023 29

4 Conclusions

In this section, the lessons learned from the highlights and lowlights of the whole research work have been documented. Additionally, we suggest some directions for potential future work based on this investigation.

4.1 Lessons Learned What Went Right 1. Being able to download evaluation copies helped in the installation and testing of the tools.

2. The required additional software necessary for the installation and configuration of the selected tools (for instance, LEIF requires VC++) was provided by our university.

3. The creation of services out of the COTS components took almost no time when the tools were used. (The user should be aware of the component and the business logic required to create a service from that component.)

4. IBM WebSphere proved to be a highly interactive tool which enhanced the usability and intelli-gibility of the feeder component (Java) and was able to generate the Web services from these components in just 4-5 minutes.

5. The Communication between the two components using the SOAP mechanism was successfully completed using the IBM WebSphere.

What Went Wrong 1. Expiration of the evaluation copies often forced us to change machine configurations and set-

tings in the development environment.

2. The COTS component was revised and is no longer freely available, thus this experiment isn’t freely repeatable.

3. No configuration management of source files is maintained due to the auto generation of the source code by the tools.

4. We could only run the application in the version of the evaluation copy that created it. Running it in a different version required extra effort and time for reconfiguration and caused data loss.

5. LEIF was unable to generate the WSDL file for the C++ Component, so it involved extra effort to find an alternative to do the same. [This was due to the incompatibility in the versioning of the source code of the Market Data Feed Component (C++).]

30 CMU/SEI-2004-TR-023

4.2 Future Directions of the Research In this section we want to highlight the Evaluation Framework’s applicability to the other model problems and tools by answering the following questions.

• Is the framework applicable to all tools and all model problems?

• How much time is needed to modify your framework when you must support multiple model problems?

• How much effort is needed in terms of searching for technologies and characterizing the model problem different ways?

4.3 Remarks While the development of the evaluation framework took more time than expected, we believe that the result is worthwhile. The framework, without change, can be used for a significant number of similar evaluations and, with minor change, could be used for a wider range of problems. Further, the evaluations contained herein demonstrate that it is possible to use the framework to distinguish be-tween tools.

The difficulties we had with the various tools suggest that, although rapid integration technologies are being widely hyped, in practice the tools still leave something to be desired. While it is possible to use the tools to integrate legacy components more efficiently than without the tools, the difficulties suggest that more work remains to be done on the tools themselves (as well, perhaps, as the target environment of Web services).

CM

U/S

EI-

2004

-TR

-023

31

Ap

pen

dix

A

Too

l Stu

die

s an

d A

nal

ysis

Lis

t of

Too

ls a

nd T

ool I

nfor

mat

ion

The

info

rmat

ion

prov

ided

her

e co

mes

dir

ectly

fro

m o

nlin

e lit

erat

ure

prov

ided

by

the

vend

ors.

# N

ame

D

escr

ipti

on

Ven

dor

Fea

ture

s P

rovi

ded

Ref

eren

ces

1.

IBM

Rat

iona

l R

apid

Dev

elop

er

IBM

Rat

iona

l Rap

id D

evel

-op

er is

a s

ingl

e, in

tegr

ated

ap

plic

atio

n de

velo

pmen

t en

viro

nmen

t tha

t com

bine

s m

odel

dri

ven

deve

lopm

ent,

arch

itect

ed R

apid

App

lica-

tion

Dev

elop

men

t (R

AD

) te

chni

ques

, and

aut

omat

ed

cons

truc

tion

to d

evel

op, i

n-te

grat

e, d

eplo

y, a

nd m

ain-

tain

hig

h-qu

alit

y, n

-tie

r bu

sine

ss a

ppli

catio

ns—

wit

hout

mos

t of

the

com

-pl

exiti

es o

f th

e un

derl

ying

te

chno

logy

pla

tfor

ms

IBM

Rat

iona

l 1.

pr

ovid

es s

ingl

e se

amle

ss e

nvir

onm

ent f

or r

apid

de

sign

, int

egra

tion,

con

stru

ctio

n, a

nd d

eplo

y-m

ent o

f bu

sine

ss a

pplic

atio

ns a

nd p

orta

ls

2.

leve

rage

s m

ains

trea

m d

evel

opm

ent s

kills

on

com

plex

n-t

ier

deve

lopm

ent p

roje

cts

3.

focu

ses

deve

lopm

ent e

ffor

ts o

n hi

gh-v

alue

bu

sine

ss r

equi

rem

ents

—no

t inf

rast

ruct

ure

co

ding

4.

acce

lera

tes

appl

icat

ion

and

port

al d

eliv

ery

thro

ugh

arch

itect

ed R

AD

tech

niqu

es s

uch

as

reve

rse-

engi

neer

ing,

rap

id u

ser

inte

rfac

e de

-si

gn, f

unct

iona

l pro

toty

ping

, aut

omat

ed n

-tie

r co

nstr

uctio

n, a

nd d

eplo

ymen

t fea

ture

s

5.

sim

plif

ies

lega

cy in

tegr

atio

n

6.

crea

tes

agil

e, s

tand

ards

-bas

ed a

pplic

atio

ns

7.

prov

ides

a v

isua

l mod

el-d

rive

n an

d po

int-

and-

clic

k en

viro

nmen

t

http

://w

ww

-30

6.ib

m.c

om/s

oftw

are/

awdt

ools

/r

apid

deve

lope

r/

http

://w

ww

-30

6.ib

m.c

om

/sof

twar

e/aw

dtoo

ls

/rap

idde

velo

per

/fea

ture

s/in

dex.

htm

l

32

CM

U/S

EI-

2004

-TR

-023

8.

prov

ides

aut

omat

ic c

onst

ruct

ion

and

hot d

e-pl

oym

ent o

f ap

plic

atio

ns a

nd p

ortl

ets

from

m

odel

s

9.

supp

orts

mai

nfra

me

and

rela

tiona

l dat

abas

e co

nnec

tivity

and

CIC

S vi

sual

and

non

-vis

ual

tran

sact

ion

inte

grat

ion

10.

XM

L m

essa

ge m

appi

ng

2.

Art

ix

Art

ix is

fam

ily

of s

ervi

ce-

orie

nted

inte

grat

ion

prod

ucts

th

at w

ill r

enov

ate

exis

ting

IT a

sset

s an

d co

nsol

idat

e le

gacy

mid

dlew

are

to r

educ

e th

e co

mpl

exit

y an

d co

st o

f IT

ope

ratio

ns. A

t the

sam

e ti

me,

it r

educ

es v

endo

r de

-pe

nden

cies

and

enh

anci

ng

futu

re I

T in

nova

tion.

Fol

-lo

win

g ar

e th

e 3

Art

ix p

rod-

ucts

.

Rel

ay

Enc

ompa

ss

Mai

nfra

me

ION

A

Tech

nolo

gies

A

rtix

Rel

ay

1.

supp

orts

a w

ide

arra

y of

mid

dlew

are

form

ats,

in

clud

ing

SOA

P an

d II

OP

2.

supp

orts

the

mos

t pop

ular

mid

dlew

are

tech

-no

logi

es, i

nclu

ding

Web

ser

vice

s,

CO

RB

A/I

IOP,

Web

Sphe

re m

q (m

qser

ies)

, bea

tu

xedo

, TIB

CO

Ren

dezv

ous

3.

exte

nds

secu

rity

, tra

nsac

tion,

and

rou

ting

fea-

ture

s of

ent

erpr

ise

mid

dlew

are

acro

ss m

iddl

e-w

are

boun

dari

es

Art

ix E

ncom

pass

1.

adds

ent

erpr

ise

feat

ures

to W

eb s

ervi

ces

inte

-gr

atio

n so

lutio

ns, i

nclu

ding

a.

secu

re W

eb s

ervi

ces

a.

st

atef

ul W

eb s

ervi

ces

b.

tr

ansa

ctio

nal W

eb s

ervi

ces

c.

m

ulti

-tra

nspo

rt W

eb s

ervi

ces

Art

ix M

ainf

ram

e

1.

high

-per

form

ance

—sh

atte

rs th

e m

isco

ncep

tion

http

://w

ww

.iona

.co

m/p

rodu

cts/

artix

/

http

://w

ww

.iona

.co

m/d

evce

nter

/art

ix/

http

://w

ww

.iona

.co

m/w

hite

pape

r.htm

#art

ix

CM

U/S

EI-

2004

-TR

-023

33

that

Web

ser

vice

s ca

nnot

be

fast

2.

auto

mat

ed—

crea

tes

new

Web

ser

vice

s in

a f

ew

min

utes

, sta

rt to

fin

ish

3.

secu

re—

inte

grat

ed w

ith S

AF

for

host

aut

hent

i-ca

tion

and

auth

oriz

atio

n

4.

stan

dard

—W

SDL

inte

rfac

e de

fini

tion

s ca

n be

us

ed b

y m

any

thir

d-pa

rty

clie

nt d

evel

opm

ent

tool

s

3.

Lig

htw

eigh

t En-

terp

rise

Int

egra

-ti

on F

ram

ewor

k (L

EIF

)

Rog

ue W

ave’

s L

EIF

is a

co

mpl

ete

best

-of-

bree

d so

lu-

tion

fra

mew

ork

for

deve

lop-

ers

need

ing

to e

nabl

e th

eir

appl

icat

ions

for

com

mun

ica-

tion

acr

oss

the

ente

rpri

se.

By

prov

idin

g m

oder

n so

lu-

tion

s ba

sed

on s

tand

ards

su

ch a

s H

TT

P (c

lient

and

se

rver

), S

impl

e O

bjec

t Ac-

cess

Pro

toco

l (SO

AP)

, Web

Se

rvic

es D

escr

iptio

n L

an-

guag

e (W

SDL

), a

nd E

xten

-si

ble

Mar

kup

Lan

guag

e (X

ML

), L

EIF

red

uces

de-

velo

pmen

t cos

ts a

nd o

ffer

s th

e ag

ility

to r

apid

ly e

volv

e ap

plic

atio

ns a

s bu

sine

ss

need

s ch

ange

. With

Rog

ue

Wav

e’s

LE

IF, i

t is

poss

ible

Rog

ue W

ave

Soft

war

e

1.

Rog

ue W

ave’

s L

EIF

incl

udes

thre

e tie

rs o

f fu

ncti

onal

ity

that

com

bine

to c

reat

e a

com

pre-

hens

ive,

pla

tfor

m-i

ndep

ende

nt f

ram

ewor

k fo

r w

orki

ng w

ith

Web

ser

vice

s an

d X

ML

in C

++:

a.

the

Dat

a T

ier—

a hi

gh-p

erfo

rman

ce C

++

de

velo

pmen

t pla

tfor

m f

or X

ML

b.

th

e N

etw

ork

Tie

r—a

com

plet

e ne

twor

king

an

d W

eb s

ervi

ces

clie

nt f

ram

ewor

k fo

r C

++

c.

th

e Se

rvic

e T

ier—

a ro

bust

C/C

++

Web

se

rvic

es h

ostin

g pl

atfo

rm

2.

The

LE

IF f

ram

ewor

k co

nsis

ts o

f a

broa

d ra

nge

of C

++

com

pone

nts

coup

led

wit

h co

de g

ener

a-to

rs f

or e

xtre

mel

y fa

st a

nd e

asy-

to-m

aint

ain

ap-

plic

atio

n de

velo

pmen

t. T

he L

EIF

cod

e ge

nera

-to

rs w

ork

toge

ther

to c

reat

e a

high

-lev

el A

PI

in

C+

+ th

at h

andl

es a

ll th

e ha

rd w

ork

of c

omm

u-ni

catio

n un

der

the

cove

rs.

3.

LE

IF a

lso

incl

udes

the

LE

IF P

roje

ct W

izar

d, a

n in

tuiti

ve g

raph

ical

use

r in

terf

ace

(GU

I) th

at

crea

tes

proj

ect f

iles

on a

ll s

uppo

rted

pla

tfor

ms

ww

w.r

ogue

wav

e.co

m/p

rodu

cts/

leif

34

CM

U/S

EI-

2004

-TR

-023

to q

uick

ly tr

ansf

orm

a

mon

olit

hic

appl

icat

ion

into

on

e th

at c

an n

ativ

ely

inte

ract

w

ith

all o

ther

par

ts o

f th

e te

chno

logy

infr

astr

uctu

re.

for

mos

t LE

IF ta

sks

and

mak

es it

eas

y to

inte

-gr

ate

LE

IF c

ode

gene

rati

on in

to a

n au

tom

ated

bu

ild p

roce

ss.

4.

PiiE

Pla

tfor

m

PiiE

is a

com

posi

te a

pplic

a-ti

on d

evel

opm

ent a

nd r

un-

tim

e en

viro

nmen

t tha

t del

iv-

ers

inte

grat

ion

to th

e de

sk-

top

by m

akin

g it

eas

y to

bu

ild a

pplic

atio

ns th

at a

re

com

pose

d fr

om m

any

back

-en

d sy

stem

s, a

nd e

xpos

ing

them

to e

nd-u

sers

in a

n ac

-ti

onab

le in

terf

ace.

The

PiiE

Pla

tfor

m c

onsi

sts

of p

rim

ary

com

pone

nts:

• Pi

iE S

mar

t Cli

ent

• Pi

iE F

usio

n Se

rver

Dig

ital

H

arbo

r P

iiE S

mar

t C

lient

1.

Dra

g an

d dr

op a

pplic

atio

n de

sign

: The

PiiE

C

onso

le B

uild

er m

akes

it e

asy

for

deve

lope

rs to

bu

ild n

ew a

pplic

atio

ns b

y us

ing

exte

nsiv

e dr

ag-

and-

drop

cap

abili

ties

to c

reat

e co

rrel

ated

be-

havi

ors.

Suc

h dr

ag-a

nd-d

rop

corr

elat

ion

is u

n-pr

eced

ente

d in

any

cli

ent/s

erve

r or

Web

dev

el-

opm

ent e

nvir

onm

ent t

oday

.

2.

Smar

t app

licat

ion

deli

very

and

man

agem

ent:

D

igit

al H

arbo

r Pi

iE p

rovi

des

over

20

Web

ser

-vi

ces

to m

anag

e ju

st-i

n-ti

me,

just

-eno

ugh

stre

amin

g of

app

licat

ions

, as

wel

l as

secu

rity

, ca

chin

g, v

ersi

onin

g, a

nd c

orre

lati

on o

f be

hav-

ior.

The

se s

ervi

ces

com

pris

e Ja

vaB

eans

and

X

ML

app

lied

at th

e in

terf

ace

leve

l.

3.

Stan

dard

s-ba

sed

appl

icat

ion

deve

lopm

ent e

nvi-

ronm

ent:

Pii

E S

mar

t Clie

nt is

des

igne

d fo

r in

-te

grat

ion

wit

h yo

ur e

xist

ing

infr

astr

uctu

re, i

n-cl

udin

g da

ta s

ourc

es a

nd a

pplic

atio

n se

rver

s.

By

prov

idin

g a

stan

dard

s-ba

sed

plat

form

for

m

anag

ing

the

appl

icat

ion

cons

umpt

ion

laye

r, Pi

iE d

ram

atic

ally

red

uces

the

tech

nica

l com

-pl

exit

y of

dev

elop

ing

appl

icat

ions

.

http

://w

ww

.dha

rbor

.com

CM

U/S

EI-

2004

-TR

-023

35

4.

Secu

rity

: Pii

E S

mar

t Clie

nt e

xten

ds r

eal-

tim

e in

tegr

atio

n w

ith e

xter

nal s

ecur

ity s

yste

ms

cru-

cial

to e

xtra

net o

pera

tions

wit

h su

ppor

t for

L

ight

wei

ght D

irec

tory

Acc

ess

Prot

ocol

(L

DA

P). P

iiE a

lso

prov

ides

aut

omat

ic s

uppo

rt

for

end-

to-e

nd e

ncry

ptio

n.

5.

Sam

ple

appl

icat

ions

: The

Pii

E S

mar

t Cli

ent

prod

uct s

uite

com

es e

quip

ped

with

out

-of-

the-

box

sam

ple

appl

icat

ions

that

pro

vide

app

lica-

tion

dev

elop

ers

and

busi

ness

ana

lyst

s w

ith e

x-is

ting

solu

tions

that

can

be

leve

rage

d fo

r cu

s-to

m b

uilt

app

lica

tions

.

6.

Col

labo

ratio

n vi

a e-

mai

l: U

sers

can

leve

rage

th

eir

exis

ting

emai

l app

lica

tions

to s

end

and

re-

ceiv

e sm

art a

pplic

atio

ns.

7. E

xten

sibl

e ID

E: P

iiE

pro

vide

s ap

plic

atio

n de

-ve

lope

rs w

ith a

n ex

tens

ible

inte

grat

ed d

evel

-op

men

t env

iron

men

t tha

t let

s de

velo

pers

plu

g in

re-

usab

le c

usto

m-b

uilt

com

pone

nts.

PiiE

Fus

ion

Serv

er

1.

One

-cli

ck d

isco

very

(A

sk):

Dig

ital

Har

bor's

Pi

iE a

llow

s us

ers

to a

sk q

uest

ions

abo

ut th

eir

info

rmat

ion

in r

eal t

ime

and

in c

onte

xt. Q

ues-

tion

s ca

n pe

rtai

n to

info

rmat

ion

abou

t rel

atio

n-sh

ips,

rol

es, r

ules

, pro

cess

es, p

eopl

e, a

nd

activ

itie

s.

2.

Con

text

ual c

olla

bora

tion

with

app

licat

ion

link-

ing

and

embe

ddin

g (A

LE

TM

): D

igit

al H

arbo

r's

36

CM

U/S

EI-

2004

-TR

-023

PiiE

lets

use

rs d

rag

and

drop

one

pie

ce o

f in

-fo

rmat

ion

onto

ano

ther

to p

rovi

de c

onte

xt f

or

colla

bora

tion.

Ent

erpr

ises

get

a “

com

mon

ope

r-at

ing

pict

ure.

”

3.

Dyn

amic

wor

kflo

w: M

ost w

orkf

low

tool

s ar

e st

and-

alon

e pa

ckag

es th

at f

ocus

on

rigi

d, p

re-

defi

ned

proc

esse

s. A

new

cla

ss o

f B

PM (

busi

-ne

ss p

roce

ss m

anag

emen

t) te

chno

logi

es is

em

ergi

ng to

inte

grat

e w

orkf

low

s ac

ross

bac

k-en

d sy

stem

s. W

hat i

s st

ill m

issi

ng, h

owev

er, i

s th

e ab

ility

to g

ive

end-

user

s co

ntro

l ove

r ho

w

flex

ibly

they

wor

k to

geth

er—

and

to d

o it

in th

e co

ntex

t of

the

appl

icat

ions

they

use

!

4.

Rat

her

than

des

ign-

tim

e w

orkf

low

, PiiE

Fus

ion

Serv

er p

rovi

des

run-

tim

e fl

exib

ility

to e

nd u

s-er

s: L

ike

a pr

ojec

t pla

nnin

g to

ol th

at is

alw

ays

live

, rat

her

than

pos

ted

on th

e w

all i

n th

e lu

nch-

room

, PiiE

's D

ynam

ic W

orkf

low

fea

ture

s al

low

en

d us

ers

to a

nnot

ate,

man

ipul

ate,

and

upd

ate

thei

r pr

oces

ses.

5.

Dra

g-an

d-D

rop

Inte

grat

ion

/ Gra

phic

al M

odel

-in

g: D

igit

al H

arbo

r’s

PiiE

mak

es in

tegr

atio

n ea

sier

by

defi

ning

a r

elat

ions

hip

mod

el o

f pe

o-pl

e, p

lace

s, a

nd th

ings

that

can

be

used

by

busi

-ne

ss a

naly

sts

to b

uild

app

licat

ions

.

6.

Cre

ate

rich

ent

ity

rela

tions

hips

acr

oss

syst

ems

by d

efin

ing

the

mea

ning

of

the

rela

tions

hip

(se-

man

tics

), n

ot ju

st th

e sy

ntax

.

a.

Map

act

ivit

ies

and

prop

ertie

s.

b.

Map

eve

nts

that

res

ult f

rom

cro

ss-s

yste

m

CM

U/S

EI-

2004

-TR

-023

37

func

tion

s.

c.

Cre

ate

or im

port

rul

es th

at c

onst

rain

the

effe

cts

of e

vent

s an

d ac

tivi

ties

on d

ata

and

peop

le.

d.

Map

to e

xter

nal a

pplic

atio

ns a

nd d

ata

sour

ces.

7.

Opt

imal

arc

hite

ctur

e: S

tand

ards

-Bas

ed D

igita

l H

arbo

r's P

iiE

is d

esig

ned

to f

it yo

ur e

xist

ing

in-

fras

truc

ture

, inc

ludi

ng d

ata

sour

ces

and

appl

ica-

tion

ser

vers

. Pii

E p

rovi

des

XM

L-b

ased

and

JD

BC

com

plia

nt d

ata

tran

sfer

and

tool

s fo

r ap

-pl

icat

ion

layo

ut a

nd d

evel

opm

ent.

8.

Ric

h us

e of

com

pone

nts

and

sem

anti

cs: E

ach

obje

ct is

“aw

are”

of

its r

elat

ions

hip

to o

ther

ob-

ject

s. E

ach

conc

ept i

n a

busi

ness

is d

escr

ibed

as

a pe

rson

, pla

ce, o

r th

ing

whi

ch h

as m

eani

ng to

ot

her

info

rmat

ion,

oth

er p

roce

sses

, and

oth

er

appl

icat

ions

.

9.

Smar

t app

licat

ion

deli

very

and

man

agem

ent:

D

igit

al H

arbo

r Pi

iE p

rovi

des

over

20

appl

ica-

tion

ser

vice

s to

man

age

just

-in-

tim

e, ju

st-

enou

gh s

trea

min

g of

app

licat

ions

and

con

tent

, se

curi

ty, c

achi

ng, v

ersi

onin

g, a

nd c

orre

latio

n of

be

havi

or.

38

CM

U/S

EI-

2004

-TR

-023

5.

SQL

Ser

ver

SQL

Ser

ver

2000

exc

eeds

de

pend

abili

ty r

equi

rem

ents

an

d pr

ovid

es in

nova

tive

ca-

pabi

litie

s th

at in

crea

se e

m-

ploy

ee e

ffec

tive

ness

, int

e-gr

ate

hete

roge

neou

s IT

ec

osys

tem

s, a

nd m

axim

ize

capi

tal a

nd o

pera

ting

budg

-et

s. S

QL

Ser

ver

2000

pro

-vi

des

the

ente

rpri

se d

ata

man

agem

ent p

latf

orm

you

r or

gani

zati

on n

eeds

to a

dapt

qu

ickl

y in

a f

ast-

chan

ging

en

viro

nmen

t.

Wit

h th

e lo

wes

t im

plem

en-

tatio

n an

d m

aint

enan

ce c

osts

in

the

indu

stry

, SQ

L S

erve

r 20

00 d

eliv

ers

rapi

d re

turn

on

you

r da

ta m

anag

emen

t in

vest

men

t. SQ

L S

erve

r 20

00 s

uppo

rts

the

rapi

d de

-ve

lopm

ent e

nter

pris

e-cl

ass

busi

ness

app

lica

tions

that

ca

n gi

ve y

our

com

pany

a

criti

cal c

ompe

titiv

e ad

van-

tage

.

Ben

chm

arke

d fo

r sc

alab

ility

, sp

eed,

and

per

form

ance

,

Mic

roso

ft

1.

easy

-to-

use

busi

ness

inte

llige

nce

tool

s

2.

self

-tun

ing

and

man

agem

ent c

apab

ilitie

s

3.

data

man

agem

ent a

pplic

atio

ns a

nd s

ervi

ces

4.

nativ

e X

ML

Sup

port

5.

upco

min

g ve

rsio

n of

SQ

L S

erve

r co

de-n

amed

“Y

ukon

” w

ill a

llow

inte

grat

ion

with

Mic

roso

ft

CL

R

6.

adva

ncem

ents

suc

h as

XQ

uery

and

a n

ativ

e X

ML

dat

a ty

pe w

ill h

elp

enab

le o

rgan

izat

ions

to

sea

mle

ssly

con

nect

inte

rnal

and

ext

erna

l sys

-te

ms

http

://w

ww

.mic

roso

ft.c

om/s

ql

CM

U/S

EI-

2004

-TR

-023

39

SQL

Ser

ver

2000

is a

ful

ly

ente

rpri

se-c

lass

dat

abas

e pr

oduc

t, pr

ovid

ing

core

sup

-po

rt f

or X

ML

and

Int

erne

t qu

erie

s.

6.

Hos

t Int

egra

tion

Serv

er 2

000

Hos

t Int

egra

tion

Serv

er

2000

ext

ends

Mic

roso

ft

Win

dow

s to

oth

er s

yste

ms

by p

rovi

ding

app

licat

ion,

da

ta, a

nd n

etw

ork

inte

gra-

tion

. Hos

t Int

egra

tion

Serv

er

lets

you

qui

ckly

ada

pt to

ne

w b

usin

ess

oppo

rtun

ities

w

hile

pre

serv

ing

exis

ting

infr

astr

uctu

re in

vest

men

ts.

Wit

h its

ent

erpr

ise-

clas

s sc

alab

ility

, per

form

ance

, and

re

liabi

lity,

Hos

t Int

egra

tion

Se

rver

can

sup

port

the

mos

t de

man

ding

bus

ines

s ne

eds.

Mic

roso

ft

1.

Ext

endi

ng c

once

pt o

f a

gate

way

: Get

com

pre-

hens

ive

Win

dow

s ap

plic

atio

n in

tegr

atio

n. H

ost

Inte

grat

ion

Serv

er 2

000

prov

ides

hos

t-to

-In

tern

et o

r ho

st-t

o-in

tran

et a

pplic

atio

n de

vel-

opm

ent c

apab

ilitie

s.

2.

Rap

id d

evel

opm

ent:

Hel

p de

velo

pers

qui

ckly

bu

ild d

istr

ibut

ed c

lient

/ser

ver

solu

tion

s th

at d

i-re

ctly

acc

ess

mai

nfra

me

or A

S/40

0 re

sour

ces

wit

hout

req

uiri

ng d

etai

led

know

ledg

e of

or

trai

ning

on

mai

nfra

me

or A

S/40

0 pr

ogra

mm

ing.

3.

Com

plet

e, s

ecur

e ac

cess

to d

ata:

Pro

vide

ob-

ject

-ori

ente

d an

d pr

ogra

mm

atic

acc

ess

to r

ela-

tion

al D

B2

data

and

fla

t file

dat

a on

mai

n-fr

ames

, AS/

400,

UN

IX, W

indo

ws

2000

, and

W

indo

ws

NT

Ser

ver

syst

ems.

Ena

ble

cli-

ent/s

erve

r-ba

sed

appl

icat

ions

to a

cces

s th

is d

ata

tran

spar

entl

y as

if d

ata

wer

e lo

cal t

o th

e se

rver

.

4.

Tra

nsac

tion

and

e-c

omm

erce

sup

port

: Int

egra

te

Mic

roso

ft T

rans

actio

n Se

rver

and

CO

M+

with

IB

M's

CIC

S or

IM

S tr

ansa

ctio

n en

viro

nmen

ts

for

drag

-and

-dro

p si

mpl

icity

to d

istr

ibut

ed,

http

://w

ww

.mic

roso

ft.c

om/h

iser

ver/

def

ault.

asp

40

CM

U/S

EI-

2004

-TR

-023

tran

sact

ion

appl

icat

ion

deve

lopm

ent,

incl

udin

g su

ppor

t for

two-

phas

ed c

omm

it b

etw

een

plat

-fo

rms.

5.

Mic

roso

ft C

omm

erce

Ser

ver

and

Biz

Talk

Se

rver

sup

port

: Int

egra

te w

ith

the

fast

est-

grow

ing

serv

er-b

ased

pla

tfor

ms

for

busi

ness

-to-

cons

umer

and

bus

ines

s-to

-bus

ines

s se

rvic

es

thro

ugh

the

Web

.

6.

Obj

ect-

orie

nted

pro

gram

min

g m

odel

: Rea

p th

e be

nefi

ts o

f ob

ject

-ori

ente

d, d

istr

ibut

ed a

pplic

a-ti

on w

ith

rapi

d de

velo

pmen

t, co

de r

euse

, sim

-pl

er m

aint

enan

ce—

pres

ervi

ng in

vest

men

t in

exis

ting

appl

icat

ions

and

dat

a.

7.

Pla

tfor

m f

or c

ompl

emen

tary

pro

duct

s: W

ork

wit

h a

soft

war

e de

velo

per

kit (

SDK

) to

pro

vide

cu

stom

ized

sol

utio

ns w

ith th

ird-

part

y

prov

ider

s.

7.

Perv

asiv

e D

ata

Junc

tion

Perv

asiv

e D

ata

Junc

tion

ad-

dres

ses

two

clos

ely

rela

ted

and

conv

ergi

ng p

robl

em

dom

ains

in tw

o im

port

ant

way

s. F

irst

, the

tech

nolo

gy

enab

les

the

rapi

d an

d re

li-ab

le c

onst

ruct

ion

of in

ter-

face

s be

twee

n tr

adin

g pa

rt-

ners

; and

, sec

ondl

y, th

e te

chno

logy

use

s th

e sa

me

inte

grat

ion

plat

form

to c

on-

nect

all

of th

e m

ajor

app

lica-

Per

vasi

ve

Solu

tion

s

Perv

asiv

e be

lieve

s th

at a

s th

e co

nver

genc

e be

twee

n B

2Bi a

nd E

AI

deve

lops

, it a

lso

seem

s th

at b

oth

inte

-gr

atio

n ch

alle

nges

can

be

serv

ed b

y a

com

mon

tool

-se

t. D

evel

opin

g, im

plem

entin

g, a

nd m

aint

aini

ng tw

o di

ffer

ent t

ools

ets

sim

ply

cost

s to

o m

uch

and

in-

volv

es r

edun

dant

sol

utio

n ar

chite

ctur

e. T

his

insi

ght

incl

udes

thre

e ba

sic

prec

epts

:

1.

Rap

id a

nd la

rge-

scal

e in

tegr

atio

n ca

n oc

cur

only

via

the

loos

ely

coup

led

exch

ange

of

busi

-ne

ss d

ocum

ents

.

2.

Dis

trib

uted

arc

hite

ctur

es w

ith in

tegr

atio

n en

-

http

://pe

rvas

ive.

co

m/e

ai/

CM

U/S

EI-

2004

-TR

-023

41

tion

s an

d da

ta s

yste

ms

wit

hin

your

ent

erpr

ise.

Thi

s tw

o-in

-one

app

roac

h to

in-

ter-

and

intr

a-en

terp

rise

in-

tegr

atio

n he

lps

keep

cos

ts

dow

n an

d gr

eatl

y si

mpl

ifie

s th

e co

mpl

exiti

es o

f bo

th

B2B

i and

EA

I.

gine

s ar

e re

quir

ed, w

hich

mea

ns th

at in

tegr

a-ti

on o

nly

real

ly o

ccur

s w

hen

all t

he “

edge

s” o

f yo

ur s

yste

ms/

appl

icat

ions

are

fin

ally

co

nnec

ted.

3.

Com

plex

bus

ines

s pr

oces

s au

tom

atio

n an

d in

-te

grat

ion

rely

on

hund

reds

and

thou

sand

s of

st

atel

ess,

ful

ly a

utom

ated

inte

grat

ion

mic

ro

flow

s.

Perv

asiv

e da

ta ju

ncti

on a

ddre

sses

the

abov

e is

sues

.

8.

Mic

roso

ft B

izTa

lk

Serv

er

Biz

Talk

Ser

ver

2002

pro

-vi

des

orga

niza

tions

wit

h th

e se

rver

, too

ls, a

dapt

ers,

and

ve

rtic

al a

ccel

erat

ors

need

ed

to in

tegr

ate

and

auto

mat

e th

eir

busi

ness

es.

Mic

roso

ft

1.

supp

ort f

or S

OA

P V

ersi

on 1

.1

2.

supp

ort f

or X

ML

, XSL

T, S

MT

P, H

TT

P, P

KI

Ver

sion

http

://w

ww

.mic

roso

ft.c

om/b

izta

lk/

9.

IBM

Web

Sphe

re

Bus

ines

s In

tegr

a-ti

on

IBM

Web

Sphe

re® is

the

lead

ing

soft

war

e pl

atfo

rm

for

e-bu

sine

ss o

n de

man

d,

deliv

erin

g a

prov

en, s

ecur

e,

and

relia

ble

soft

war

e

port

folio

.

IBM

1.

IB

M W

ebSp

here

Bus

ines

s In

tegr

atio

n C

onne

ct

V4.

2 al

low

s yo

u to

con

nect

to a

nd in

tegr

ate

wit

h co

mm

unit

ies

of tr

adin

g pa

rtne

rs.

2.

supp

ort f

or X

ML

, mul

tipl

e se

curi

ty s

tand

ards

3.

IBM

Web

Sphe

re B

usin

ess

Inte

grat

ion

Serv

er

allo

ws

you

to in

tegr

ate

busi

ness

pro

cess

es.

http

://w

ww

-30

6.ib

m.c

om

/sof

twar

e/in

fo1

/web

sphe

re

/ind

ex.js

p?ta

b=hi

ghli

ghts

http

://w

ww

-10

6.ib

m.c

om

/dev

elop

erw

orks

/w

ebsp

here

/zon

es

/bus

ines

sint

egra

tion

42

CM

U/S

EI-

2004

-TR

-023

10.

Ens

embl

e E

nsem

ble

is a

com

preh

en-

sive

app

lica

tion

inte

grat

ion

plat

form

that

ena

bles

exc

ep-

tion

ally

fas

t int

egra

tion

and

ex

trem

ely

rapi

d de

velo

p-m

ent o

f co

mpo

site

app

lica-

tion

s. E

nsem

ble

exce

ls a

t bu

ildin

g ne

w s

trat

egic

bus

i-ne

ss s

olut

ions

that

leve

rage

th

e fu

ncti

onal

ity

of e

xist

ing

appl

icat

ions

, orc

hest

rate

ne

w b

usin

ess

proc

esse

s, a

nd

inte

grat

e da

ta a

cros

s th

e

ente

rpri

se.

Inte

rSys

tem

1.

Fu

ll sp

ectr

um in

tegr

atio

n an

d de

velo

pmen

t: E

n-se

mbl

e pr

ovid

es a

uni

fied

gra

phic

al-,

XM

L-,

and

code

-bas

ed d

evel

opm

ent e

nvir

onm

ent f

or b

uild

-in

g cu

stom

ada

pter

s, o

rche

stra

ting

busi

ness

pro

c-es

ses,

and

cre

atin

g co

mpo

site

app

licat

ions

.

2.

Uni

vers

al s

ervi

ce a

rchi

tect

ure:

Ens

embl

e pr

o-vi

des

a co

nsis

tent

and

eff

icie

nt o

bjec

t rep

rese

n-ta

tion

of d

iffe

rent

pro

gram

min

g m

odel

s an

d da

ta f

orm

ats.

3.

Pers

iste

nt o

bjec

t eng

ine:

At i

ts c

ore,

Ens

embl

e us

es a

hig

h-pe

rfor

man

ce d

istr

ibut

ed o

bjec

t da-

taba

se f

or m

anag

ing

and

stor

ing

all m

etad

ata,

m

essa

ges,

and

pro

cess

sta

te in

form

atio

n.

4.

Cus

tom

izab

le e

nd-t

o-en

d m

anag

emen

t and

m

onit

orin

g: E

nsem

ble

prov

ides

hig

hly-

cust

omiz

able

and

ext

ensi

ble

mon

itor

ing

and

man

agem

ent f

acili

ties

that

are

tigh

tly in

tegr

ated

w

ith

the

mod

elin

g an

d de

velo

pmen

t too

ls.

http

://w

ww

.inte

rsys

tem

s.co

m

/ens

embl

e /t

echn

olog

y /i

ndex

.htm

l

11.

JBos

s A

pplic

atio

n Se

rver

JB

oss

is a

Jav

a-ba

sed

open

so

urce

app

licat

ion

serv

er

wit

h su

ppor

t for

J2E

E. T

his

enab

les

rapi

d po

rtin

g of

J2

EE

-bas

ed a

pplic

atio

ns to

th

e fr

ee J

Bos

s se

rver

.

JBos

s su

ppor

ts b

oth

the

con-

stru

ctio

n of

Jav

a-ba

sed

Web

Se

rvic

es a

s w

ell a

s th

e in

te-

grat

ion

of p

ossi

bly

non-

Java

JBos

s G

roup

1.

fu

ll m

icro

kern

el a

ppro

ach

base

d on

Jav

a M

an-

agem

ent e

Xte

nsio

ns (

JMX

)

2.

fully

hot

-dep

loya

ble

and

cycl

eabl

e se

rvic

e la

yer

wit

h Se

rvic

e A

rchi

ve (

SAR

) fo

rmat

3.

fully

aut

omat

ed a

nd n

et-b

ased

inst

alla

tion

wit

h ho

t-de

ploy

of

appl

icat

ions

4.

full

J2E

E 1

.3 s

uppo

rt (E

JB, J

CA

, JSP

, JM

X,

HT

TP,

etc

.)

5.

full

secu

rity

impl

emen

tatio

n an

d JA

AS

inte

gra-

tion

ww

w.jb

oss.

org

CM

U/S

EI-

2004

-TR

-023

43

base

d ex

tern

al W

eb S

ervi

ces

thro

ugh

the

JBos

s.N

et e

x-te

nsio

n. A

spe

cial

foc

us is

pl

aced

on

patte

rns

and

tool

s to

exp

ose

J2E

E™

-bas

ed

logi

c, s

uch

as s

essi

on b

eans

an

d en

tity

bean

s.

6.

full

clus

teri

ng o

f an

y Ja

va o

bjec

ts (

EJB

, HT

TP)

7.

grou

nd-b

reak

ing

Asp

ect-

Ori

ente

d Pr

ogra

mm

ing

(AO

P).

44

CM

U/S

EI-

2004

-TR

-023

CM

U/S

EI-

2004

-TR

-023

45

Ap

pen

dix

B

Too

l Eva

luat

ion

Rep

ort

s

IBM

Web

Sphe

re

Cat

egor

y N

on-T

echn

ical

P

aram

eter

s V

alue

(I

n Sc

ale

of 1

to 1

0)

Rat

iona

le

Bus

ines

s

Cos

ts

Ben

efit

s

Ris

k A

naly

sis

7 C

ost

Mar

ket P

rice

: $1,

000

and

$3,5

00

Ben

efit

s

RO

I

Goo

d R

OI

wit

h th

e to

ol w

hich

is c

lear

fro

m th

e su

cces

s st

orie

s po

sted

on

the

Web

site

.

Lif

espa

n of

the

tool

8.

Cre

ate,

bui

ld, t

est,

depl

oy a

nd p

ublis

h W

eb s

ervi

ces

with

sup

port

for

Uni

vers

al D

escr

iptio

n,

Dis

cove

ry a

nd I

nteg

ratio

n (U

DD

I), V

ersi

on 2

; Sim

ple

Obj

ect A

cces

s Pr

otoc

ol (

SOA

P); W

eb

Serv

ices

Des

crip

tion

Lan

guag

e (W

SDL

); a

nd W

eb S

ervi

ces

Insp

ectio

n L

angu

age

(WSI

L).

Cre

-at

e, v

alid

ate

and

dete

ct W

S-I

Bas

ic P

rofi

le, V

ersi

on 1

.0-

com

plia

nt W

eb s

ervi

ces

for

grea

ter

in-

tero

pera

bilit

y be

twee

n se

rvic

es.

9.

Cre

ate,

test

, and

dep

loy

J2E

E a

pplic

atio

ns to

BE

A W

ebL

ogic

Ser

ver V

6.1

and

V7.

0 us

ing

the

Dep

loym

ent T

oolk

it f

or W

ebSp

here

Stu

dio,

Web

Log

ic E

ditio

n.

10.

Qui

ckly

bui

ld r

ich,

inte

ract

ive

user

inte

rfac

es f

or W

eb a

pplic

atio

ns u

sing

reu

sabl

e Ja

vaSe

rver

Fa

ces

(JSF

)

Ris

k A

naly

sis

1.

Cha

nge

Freq

uenc

y

L

ess

wit

h tw

o re

leas

es in

a y

ear.

2.

Up-

grad

atio

n

46

CM

U/S

EI-

2004

-TR

-023

a.

Eas

y up

-gra

dati

on to

new

rel

ease

s. (

e.g.

, can

eas

ily u

pgra

de th

e W

ebSp

here

Stu

dio

from

5.

1 to

5.1

.1)

Eva

luat

ion

Spec

ific

Pr

ojec

t Spe

cifi

c an

d In

depe

nden

t Fa

ctor

s

Com

pari

son

with

O

ther

Too

ls

7 P

roje

ct S

peci

fic

and

Inde

pend

ent F

acto

rs

1.

prov

ides

com

plet

e de

velo

pmen

t env

iron

men

t for

J2E

E, J

ava,

Web

Ser

vice

s, X

ML

and

Web

2.

prov

ides

Tea

m D

evel

opm

ent e

nvir

onm

ent

3.

prov

ides

Ser

ver

tool

s fo

r te

stin

g

4.

supp

orts

alm

ost 1

0 la

ngua

ges

Com

pari

son

wit

h O

ther

Too

ls

Web

Sphe

re 4

.0 A

dvan

ced

Edi

tion

cost

s ov

er 1

0 tim

es m

ore

than

Mic

roso

ft .N

ET.

For

Web

Ser

vice

Dev

elop

men

t:

IBM

Web

Sphe

re 4

.0 r

equi

res

145%

mor

e m

anua

l cod

ing

than

Vis

ual S

tudi

o .N

ET.

The

fol

low

ing

link

giv

es m

ore

info

rmat

ion

on c

ompa

rati

ve a

naly

sis

of I

BM

Web

Sphe

re w

ith o

ther

to

ols:

http

://w

ww

.got

dotn

et.c

om/t

eam

/com

pare

/web

serv

icec

ompa

re.a

spx

Ext

erna

l R

efer

ence

M

arke

t Aw

aren

ess

8 M

arke

t A

war

enes

s

Tool

is p

opul

ar a

mon

g sm

all-

and

med

ium

-siz

ed o

rgan

izat

ions

and

ther

e ar

e al

mos

t 20

succ

ess

sto-

ries

for

this

tool

list

ed o

n th

e IB

M W

eb s

ite.

http

://w

ww

-30

6.ib

m.c

om/s

oftw

are/

succ

ess/

cssd

b.ns

f/la

test

stor

iesV

W?O

penV

iew

&C

ount

=10

&R

estr

ictT

oCat

egor

y=st

udio

appd

ev

CM

U/S

EI-

2004

-TR

-023

47

Ven

dor

Su

ppor

t

Qua

lity

and

Cos

t of

Ven

dor

Supp

ort

Acc

ess

to I

nter

nal

Tool

Inf

orm

atio

n

Ver

sion

Cho

ice

6 Q

ualit

y an

d C

ost

of V

endo

r Su

ppor

t

Tech

nica

l sup

port

fro

m th

e ve

ndor

is v

ery

good

. (O

ne u

ser

was

not

abl

e to

con

figu

re th

e J2

EE

Uni

t te

st s

erve

r. I

post

ed th

is p

robl

em to

the

tech

nica

l sup

port

and

got

rep

ly th

e ne

xt d

ay.)

Acc

ess

to I

nter

nal T

ool I

nfor

mat

ion

No

acce

ss to

arc

hite

ctur

e an

d de

sign

asp

ects

of

the

tool

Ver

sion

Cho

ice

Web

Sphe

re A

ppli

catio

n D

evel

opm

ent S

tudi

o 5.

1 be

st s

uit t

he r

equi

rem

ents

.

Too

l-Sp

ecif

ic

Inte

grab

ility

and

C

ompa

tibi

lity

Tailo

rabi

lity

Dom

ain

Tool

Int

egri

ty a

nd

Stan

dard

s

Skil

ls b

ase

and

Use

r C

apab

ility

Dat

a C

onfi

gura

-ti

on C

ontr

ol

Tool

Mod

ular

ity

Res

pons

e T

ime

Use

r In

terf

ace

8 In

tegr

abili

ty a

nd C

ompa

tibi

lity

1.

built

on

Ecl

ipse

and

can

be

cust

omiz

ed a

nd e

xten

ded

with

a w

ide

rang

e of

IB

M P

rodu

cts

and

part

ner

and

Ecl

ipse

bas

ed p

lug-

ins.

2.

inte

grat

ion

with

sof

twar

e co

nfig

urat

ion

man

agem

ent t

ools

3.

supp

orts

Win

dow

s an

d L

inux

pla

tfor

m

Tai

lora

bilit

y

Tool

can

be

tailo

red

to th

e us

er r

equi

rem

ents

usi

ng b

uilt

-in

wiz

ards

and

plu

g-in

s.

Dom

ain

Prov

ides

J2E

E, J

ava,

Web

Ser

vice

s, X

ML

, and

Web

Dev

elop

men

t Env

iron

men

t and

als

o pr

ovid

es

rela

tiona

l dat

abas

e to

ols.

Too

l Int

egri

ty a

nd S

tand

ards

1.

Su

ppor

t for

Uni

vers

al D

escr

ipti

on, D

isco

very

and

Int

egra

tion

(U

DD

I), V

ersi

on 2

; SO

AP;

W

SDL

; and

Web

Ser

vice

s In

spec

tion

Lan

guag

e (W

SIL

), J

2EE

2.0

and

3.0

, UM

L, S

QL

J, e

tc.

2.

Tool

is v

ery

relia

ble;

it p

rom

pts

the

erro

r m

essa

ges

as w

hen

user

inpu

ts a

ny v

alue

.

3.

Ven

dor

Mat

urit

y is

qui

te h

igh.

48

CM

U/S

EI-

2004

-TR

-023

Doc

umen

tatio

n

Dat

a Po

rtab

ility

Skill

s B

ase

and

Use

r C

apab

ility

Nor

mal

use

r w

ith

know

ledg

e of

J2E

E, W

eb S

ervi

ces

and

Web

Dev

elop

men

t can

ope

rate

this

tool

. T

he to

ol p

rovi

des

diff

eren

t dev

elop

men

t vie

ws

(UI)

kno

wn

as p

er p

ersp

ecti

ve a

nd p

er te

rmin

olog

y fo

r ea

ch d

evel

opm

ent e

nvir

onm

ent.

For

Exa

mpl

e in

J2E

E p

ersp

ecti

ve th

e to

ol d

ispl

ays

the

J2E

E A

r-ch

itect

ure

Com

pone

nts

on th

e le

ft n

avig

atio

nal b

ar. (

In o

ne c

ase

it to

ok a

wee

k to

get

acq

uain

ted

wit

h th

e ba

sic

feat

ures

pro

vide

d by

the

tool

and

to d

evel

op a

sam

ple

J2E

E a

pplic

atio

n w

hich

add

s M

P3 s

ong

info

rmat

ion

to th

e C

loud

scap

e da

taba

se u

sing

the

Web

inte

rfac

e in

thre

e ho

urs.

)

Con

figu

rati

on C

ontr

ol

App

lica

tion

Dev

elop

er p

rovi

des

mon

itor

ing

and

prof

iling

tool

s th

at f

eatu

res

cust

omiz

able

vie

ws

and

logs

ena

blin

g yo

u to

rec

ogni

ze, i

sola

te, a

nd f

ix p

erfo

rman

ce p

robl

ems

earl

y in

the

deve

lopm

ent c

y-cl

e, s

uppo

rt f

or o

bjec

t lev

el p

rofi

ling,

ana

lysi

s of

Web

Sphe

re A

pplic

atio

n Se

rver

act

ivit

y lo

gs, a

nd

inte

ract

ion

with

the

sym

ptom

dat

abas

e an

d L

og a

nd T

race

Ana

lyze

r.

Too

l Mod

ular

ity

1.

built

on

Ecl

ipse

2.

pr

ovid

es c

ompr

ehen

sive

dev

elop

men

t env

iron

men

t thr

ough

vis

ual t

ools

, tem

plat

es,

code

gen

erat

ion

utili

ties

Use

r In

terf

ace

GU

I ba

sed

deve

lopm

ent m

akes

the

tool

ver

y us

er f

rien

dly.

Hel

pful

err

or m

essa

ges

help

in lo

catin

g th

e ca

use

of e

rror

. R

espo

nse

Tim

e

Perf

orm

ance

of

the

tool

is g

ood.

Use

r In

terf

ace

Doc

umen

tati

on

Tool

is b

undl

ed w

ith s

uffi

cien

t doc

umen

tatio

n. T

he o

ffic

ial W

eb s

ite

prov

ides

tech

nica

l inf

orm

atio

n an

d ha

s lo

ts o

f di

scus

sion

for

ums

to e

lici

t any

kin

d of

info

rmat

ion.

CM

U/S

EI-

2004

-TR

-023

49

Dat

a P

orta

bilit

y

Dat

a ge

nera

ted

by th

e to

ol is

mos

tly

in te

xt f

orm

at w

hich

can

be

view

ed b

y an

y te

xt e

dito

r.

Em

plac

emen

t In

tegr

atio

n w

ith

Oth

er T

ools

7

Inte

grat

ion

wit

h O

ther

Too

ls

1.

Inte

grat

ion

with

sev

eral

Sof

twar

e C

onfi

gura

tion

Man

agem

ent s

olut

ions

; pro

vide

s fl

exib

ility

in

asse

t man

agem

ent a

nd te

am d

evel

opm

ent

2.

Tool

use

s V

ersi

on C

ontr

ol M

anag

emen

t int

erfa

ce p

rovi

ded

by E

clip

se V

2.1

and

an a

dapt

or f

or

IBM

Cle

arC

ase

LT a

nd f

or C

VS.

IBM

Rat

iona

l Rap

id D

evel

oper

Cat

egor

y N

on-T

echn

ical

P

aram

eter

s V

alue

(I

n Sc

ale

of 1

to 1

0)

Rat

iona

le

Bus

ines

s

Cos

ts

Ben

efit

s

Ris

k A

naly

sis

5 C

ost

Mar

ket P

rice

: $ 5

995

per

vers

ion

Ben

efit

s 1.

T

he to

ol h

as e

asy-

to-u

se m

odel

ing

capa

bilit

ies.

2.

Lif

espa

n of

the

tool

a.

The

tool

sup

port

s th

e de

sign

, int

egra

tion,

dep

loym

ent,

cons

truc

tion

of b

usin

ess

appl

ica-

tion

s. I

t aut

omat

es m

any

task

s so

that

dev

elop

ers

focu

s on

hig

hest

val

ue a

reas

. b.

T

he R

apid

Dev

elop

er R

untim

e A

PI

allo

ws

appl

icat

ion

deve

lope

rs to

cre

ate

this

bus

ines

s lo

gic

usin

g a

stan

dard

ized

inte

rfac

e th

at w

orks

und

er a

ll su

ppor

ted

tech

nolo

gies

. c.

R

apid

Dev

elop

er s

uppo

rts

the

desi

gn a

nd c

onst

ruct

ion

of p

ages

, mes

sage

s, c

ompo

nent

s,

and

Web

Ser

vice

s us

ing

eith

er J

2EE

or

MSD

NA

Tec

hnol

ogie

s.

50

CM

U/S

EI-

2004

-TR

-023

Ris

k A

naly

sis

1.

The

tool

is h

ighl

y de

pend

ent o

n R

atio

nal T

ools

, so

the

appl

icat

ion

mus

t be

mod

eled

by

“Ra-

tion

al”

met

hodo

logy

.

Eva

luat

ion

Spec

ific

Pr

ojec

t Spe

cifi

c an

d In

depe

nden

t Fa

ctor

s

Com

pari

son

with

O

ther

Too

ls

5 P

roje

ct S

peci

fic

and

Inde

pend

ent F

acto

rs

The

tool

pro

vide

s “a

rchi

tect

ure-

cent

ric”

dev

elop

men

t env

iron

men

t whi

ch o

nly

depi

cts

a “t

ier-

base

d” v

iew

of

the

entir

e ap

plic

atio

n th

us r

estr

ictin

g th

e us

er to

dev

elop

tier

-bas

ed a

pplic

atio

ns.

Com

pari

son

wit

h O

ther

Too

ls

Mos

t of

the

mod

elin

g fe

atur

es a

re s

ame

as th

ose

in R

atio

nal S

uite

of T

ools

.

Ext

erna

l R

efer

ence

Mar

ket A

war

enes

s

6

Mar

ket

Aw

aren

ess

The

tool

is p

opul

ar a

mon

g di

ffer

ent s

ized

ent

erpr

ises

and

ther

e ar

e su

cces

s st

orie

s w

hich

are

list

ed

on th

e IB

M W

eb s

ite.

Ven

dor

Su

ppor

t

Qua

lity

and

Cos

t of

Ven

dor

Supp

ort

Acc

ess

to I

nter

nal

Tool

Inf

orm

atio

n

Ver

sion

Cho

ice

6 Q

ualit

y an

d C

ost

of V

endo

r Su

ppor

t

Ven

dor

Supp

ort a

nd S

uppo

rtin

g D

ocum

enta

tion

are

reas

onab

le b

ut n

ot te

chni

cally

sou

nd.

Acc

ess

to I

nter

nal T

ool I

nfor

mat

ion

No

acce

ss to

Arc

hite

ctur

e an

d D

esig

n as

pect

s of

the

tool

is a

vaila

ble.

Ver

sion

Cho

ice

Rel

ease

200

3.06

.00.

105.

is s

uita

ble.

Too

l Sp

ecif

ic

Inte

grab

ility

and

C

ompa

tibi

lity

6 In

tegr

abili

ty a

nd C

ompa

tibi

lity

Can

be

inte

grat

ed w

ith R

atio

nal C

lear

Cas

e, R

ose

Dia

gram

s, it

is a

lso

com

pati

ble

wit

h IB

M W

eb-

CM

U/S

EI-

2004

-TR

-023

51

Tailo

r-ab

ility

Dom

ain

Tool

Int

egri

ty a

nd

Stan

dard

s

Skil

ls B

ase

and

Use

r C

apab

ility

Dat

a C

onfi

gura

tion

Con

trol

Tool

Mod

ular

ity

Res

pons

e T

ime

Use

r In

terf

ace

Doc

umen

tatio

n D

ata

Port

abili

ty

Sphe

re A

pplic

atio

n Se

rver

, Ora

cle

9.1

App

licat

ion

Serv

er. H

owev

er, t

here

is a

n ov

er-d

epen

denc

y on

IB

M T

ools

.

Tai

lor-

abili

ty

Tool

can

be

tailo

red

to th

e us

er r

equi

rem

ents

usi

ng te

mpl

ates

and

pat

tern

s.

Dom

ain

Prov

ides

J2E

E, J

ava,

Web

Ser

vice

s, X

ML

, Leg

acy

Serv

ices

, and

Web

Dev

elop

men

t Env

iron

men

t; al

so p

rovi

des

rela

tiona

l dat

abas

e to

ols.

Too

l Int

egri

ty a

nd S

tand

ards

1.

T

he to

ol s

uppo

rts

Uni

vers

al D

escr

iptio

n, D

isco

very

and

Int

egra

tion

(UD

DI)

, Ver

sion

2;

Sim

ple

Obj

ect A

cces

s Pr

otoc

ol (

SOA

P); W

eb S

ervi

ces

Des

crip

tion

Lan

guag

e (W

SDL

); a

nd

Web

Ser

vice

s In

spec

tion

Lan

guag

e (W

SIL

), J

2EE

2.0

and

3.0

, UM

L, S

QL

J, e

tc.

2.

Tool

is r

elia

ble

and

prov

ides

goo

d ex

cept

ion-

hand

ling

fac

ilitie

s.

Skill

s B

ase

and

Use

r C

apab

ility

Any

use

r w

ith k

now

ledg

e of

J2E

E, W

eb S

ervi

ces,

or W

eb D

evel

opm

ent c

an o

pera

te th

is to

ol. T

he

tool

pro

vide

s di

ffer

ent “

arch

itect

s” a

s on

e go

es a

bout

des

igni

ng th

e ap

plic

atio

n. H

owev

er, t

he

know

ledg

e of

lega

cy d

atab

ases

is a

goo

d ad

d-on

in c

ase

the

tool

is to

be

used

for

lega

cy in

tegr

a-ti

on p

urpo

ses.

Con

figu

rati

on C

ontr

ol

The

tool

pro

vide

s fu

nctio

nalit

y fo

r in

terf

acin

g to

a v

ersi

on-c

ontr

ol s

yste

m b

ut d

oes

not h

ave

one

built

in.

Too

l Mod

ular

ity

Prov

ides

com

preh

ensi

ve d

evel

opm

ent e

nvir

onm

ent t

hrou

gh te

mpl

ates

, cod

e ge

nera

tion

utili

ties

, pa

ttern

s, a

nd v

iew

s

Use

r In

terf

ace

52

CM

U/S

EI-

2004

-TR

-023

The

GU

I-ba

sed

func

tiona

lity

is e

asy

to u

se—

spec

ific

ally

the

“tab

s” to

des

ign

the

appl

icat

ion.

Res

pons

e T

ime

Perf

orm

ance

of

the

tool

is r

easo

nabl

y ok

.

Use

r In

terf

ace

Doc

umen

tati

on

The

tool

is b

undl

ed w

ith s

uffi

cien

t doc

umen

tatio

n. T

he I

BM

Web

site

pro

vide

s a

good

res

ourc

e fo

r w

hite

pap

ers,

dem

os, e

valu

atio

n gu

ides

.

Em

plac

emen

t In

tegr

atio

n w

ith

Oth

er T

ools

7

Inte

grat

ion

wit

h O

ther

Too

ls

Inte

grat

es w

ith o

ther

non

-Rat

iona

l-ba

sed

Sour

ce C

ontr

ol S

yste

ms,

but

doe

s no

t int

egra

te w

ith

a w

ide

vari

ety

of a

pplic

atio

ns

LE

IF

Cat

egor

y N

on-T

echn

ical

P

aram

eter

s V

alue

(I

n Sc

ale

of 1

to 1

0)

Rat

iona

le

Bus

ines

s

Cos

ts

Ben

efit

s

Ris

k A

naly

sis

5 C

ost

LE

IF D

evel

oper

Net

wor

k: L

icen

se +

Bas

ic S

uppo

rt

$1,4

95.0

0

LE

IF D

evel

oper

Net

wor

k: L

icen

se +

Pre

miu

m

Supp

ort

$1,7

90.0

0 B

enef

its

Bus

ines

s

CM

U/S

EI-

2004

-TR

-023

53

Eas

e of

use

and

com

para

tivel

y sh

orte

r le

arni

ng c

urve

for

a p

erso

n w

ith d

omai

n kn

owle

dge

RO

I is

muc

h hi

gher

in c

ases

of

appl

icat

ions

dev

elop

ed u

sing

C+

+.

Pro

duct

1.

a

rapi

d de

velo

pmen

t too

l for

cre

atin

g Se

rvic

e or

ient

ed a

pplic

atio

ns in

volv

ing

C+

+

2.

crea

tes

Web

ser

vice

s us

ing

the

exis

ting

lega

cy a

pplic

atio

ns w

itho

ut n

eed

for

mod

ific

atio

n to

th

e ex

istin

g bu

sine

ss lo

gic,

ther

eby

redu

cing

tim

e sp

ent i

n re

wri

ting

the

code

for

the

sam

e

3.

addr

esse

s sc

alab

ility

and

hig

h-pe

rfor

man

ce a

ttrib

utes

of

C+

+ a

pplic

atio

ns.

Ris

k A

naly

sis

App

licat

ion

App

lyin

g se

rvic

e-or

ient

ed d

evel

opm

ent a

nd W

eb s

ervi

ces

only

for

C++

App

lica

tions

Cha

nge

Fre

quen

cy

The

cha

nge

freq

uenc

y is

hig

h si

nce

it is

und

er th

e de

bugg

ing

and

fixi

ng s

tage

of

the

prod

uct.

Al-

thou

gh th

e L

EIF

Ver

sion

2.0

is a

sta

ble

one,

bas

ed o

n w

hich

up-

grad

atio

n ca

n be

don

e, th

e ch

ange

fr

eque

ncy

may

be

high

.

Eva

luat

ion

Spec

ific

Pr

ojec

t Spe

cifi

c an

d In

depe

nden

t Fa

ctor

s

Com

pari

son

with

O

ther

Too

ls

6 P

roje

ct S

peci

fic

and

Inde

pend

ent F

acto

rs

For

the

proj

ect t

hat w

e ha

ve s

elec

ted,

the

“Tra

ding

Bon

d Sy

stem

,” L

EIF

is u

sed

for

conv

ertin

g th

e st

and-

alon

e C

++

app

licat

ion

into

a c

lient

ser

ver

appl

icat

ion.

The

cor

e C

++

fun

ctio

nalit

y is

ex

pose

d th

roug

h a

Web

ser

vice

. It w

as s

peci

fica

lly

desi

gned

to b

ridg

e th

e C

++ a

nd J

ava

gap

and

repr

esen

ted

the

leas

t am

ount

of

wor

k.

Com

pari

son

wit

h O

ther

Too

ls

The

re is

not

muc

h co

mpe

titio

n or

com

pari

son

to b

e m

ade.

sin

ce L

EIF

is o

ne a

mon

g th

e ve

ry f

ew

serv

ice-

orie

nted

rap

id d

evel

opm

ent t

ools

for

C++

app

licat

ions

. The

oth

er to

ol th

at s

uppo

rts

the

sam

e is

IO

NA

Art

ix; b

ut th

e co

mpl

exity

and

the

lear

ning

cur

ve in

volv

ed in

IO

NA

Art

ix a

re

54

CM

U/S

EI-

2004

-TR

-023

grea

ter

than

that

req

uire

d fo

r L

EIF

.

Ext

erna

l R

efer

ence

M

arke

t Aw

aren

ess

6 M

arke

t A

war

enes

s

Rog

ue W

ave

Soft

war

e In

c. is

a le

adin

g pr

ovid

er o

f re

usab

le s

oftw

are

com

pone

nts

and

serv

ices

for

ap

plic

atio

n de

velo

pmen

t. Si

nce

Dec

embe

r 19

, 200

3, it

has

bee

n a

subs

idia

ry o

f Q

uova

dx I

nc.

The

fol

low

ing

hype

rlin

ks a

cces

s th

e va

riou

s pr

ess

rele

ases

, art

icle

s th

at h

ighl

ight

ed th

e pr

oduc

t se

rvic

es o

f R

ogue

Wav

e, a

nd th

e aw

ards

gra

nted

.

http

://w

ww

.rog

uew

ave.

com

/cor

p/pr

ess/

pres

srel

/

http

://w

ww

.rog

uew

ave.

com

/cor

p/pr

ess/

arti

cles

/

http

://w

ww

.rog

uew

ave.

com

/cor

p/pr

ess/

awar

ds/

Tho

ugh

the

com

pany

enj

oys

sign

ific

ant m

arke

t aw

aren

ess,

the

prod

uct i

s st

ill u

nder

exp

erim

ent.

It d

oes

prov

ide

good

cas

e st

udie

s th

at a

re p

rove

n su

cces

s st

orie

s fo

r L

EIF

.

http

://w

ww

.rog

uew

ave.

com

/pro

duct

s/le

if/P

utna

mcs

.pdf

Ven

dor

Su

ppor

t

Qua

lity

and

Cos

t of

Ven

dor

Supp

ort

Acc

ess

to I

nter

nal

8 Q

ualit

y an

d C

ost

of V

endo

r Su

ppor

t

Use

rs c

an o

btai

n te

chni

cal s

uppo

rt b

y

1.

cont

actin

g te

chni

cal s

uppo

rt

2.

usin

g th

e kn

owle

dge

base

CM

U/S

EI-

2004

-TR

-023

55

Tool

Inf

orm

atio

n

Ver

sion

Cho

ice

3.

acce

ssin

g va

riou

s su

ppor

t pro

gram

s

4.

read

ing

prod

uct d

ocum

enta

tion

“The

res

pons

e ti

me

or th

e tu

rnar

ound

tim

e of

the

LE

IF to

ol v

endo

r su

ppor

t was

one

day

whe

n I

face

d th

e pr

oble

m w

ith

the

gene

ratio

n of

a p

roje

ct.”

http

://w

ww

.rog

uew

ave.

com

/sup

port

/

Acc

ess

to I

nter

nal T

ool I

nfor

mat

ion

The

use

r gu

ide

of th

e L

EIF

Fra

mew

ork

give

s de

taile

d in

form

atio

n ab

out L

EIF

arc

hite

ctur

e w

ith

its

thre

e tie

rs a

nd th

eir

bene

fits

. It a

lso

talk

s ab

out h

ow th

e to

ol is

act

uall

y pr

oces

sing

the

serv

ice-

orie

nted

app

licat

ions

Ver

sion

Cho

ice

Rog

ue W

ave

issu

es th

ree

type

s of

LE

IF p

rodu

ct r

elea

ses:

1.

Mai

nten

ance

(x.

y.z)

rel

ease

s pr

ovid

e bu

g fi

xes

and/

or p

latf

orm

upd

ates

.

2.

Min

or (x

.y.0

) re

leas

es p

rovi

de b

ug f

ixes

, pla

tfor

m u

pdat

es, a

nd m

inor

pro

duct

enh

ance

-m

ents

.

3.

Maj

or (x

.0.0

) re

leas

es p

rovi

de b

ugs

fixe

s, p

latf

orm

upd

ates

, and

maj

or p

rodu

ct e

nhan

ce-

men

ts o

r ne

w f

eatu

res.

The

cur

rent

maj

or v

ersi

on o

r re

leas

e of

LE

IF is

2.0

, whi

ch p

rovi

des

high

ben

efit

s in

term

s of

per

-fo

rman

ce, i

nteg

ratio

n an

d ex

tens

ibili

ty.

http

://w

ww

.rog

uew

ave.

com

/pro

duct

s/le

if/w

hats

new

.cfm

Too

l Sp

ecif

ic

Inte

grab

ility

and

C

ompa

tibi

lity

8 In

tegr

abili

ty a

nd C

ompa

tibi

lity

• B

road

mes

sagi

ng p

atte

rn s

uppo

rt: C

hoos

e th

e ap

prop

riat

e m

essa

ging

pat

tern

fro

m r

e-qu

est/r

espo

nse,

asy

nchr

onou

s (I

OU

), o

ne-w

ay, s

erve

r-si

de n

otif

icat

ion,

and

ser

ver-

side

sol

icit

56

CM

U/S

EI-

2004

-TR

-023

Tailo

r-ab

ility

Dom

ain

Tool

Int

egri

ty a

nd

Stan

dard

s

Skil

ls b

ase

and

Use

r C

apab

ility

Dat

a C

onfi

gura

tion

Con

trol

Tool

Mod

ular

ity

Res

pons

e T

ime

Use

r In

terf

ace

Doc

umen

tatio

n

Dat

a Po

rtab

ility

resp

onse

. Clie

nts

can

liste

n fo

r se

rver

-ini

tiate

d ev

ents

.

• IB

M W

ebSp

here

MQ

™ tr

ansp

ort:

Get

sup

erio

r re

liabi

lity

and

inte

grat

ion

by c

onne

ctin

g L

EIF

se

rvic

es d

irec

tly to

the

mes

sage

que

ue—

no n

eed

for

com

plic

ated

(an

d sl

ow)

brid

ges.

• E

xhau

stiv

e in

tero

pera

bilit

y te

stin

g: P

artic

ipat

ion

in g

roup

s lik

e SO

AP

Bui

lder

s en

sure

s th

at

LE

IF s

ervi

ces

can

be c

oupl

ed w

ith s

ervi

ces

expo

sed

usin

g ot

her

tech

nolo

gies

.

Tai

lora

bilit

y

Bel

ow a

re s

ome

of th

e ch

arac

teri

stic

s of

LE

IF th

at m

akes

the

prod

uct t

ailo

rabl

e.

• E

asily

wri

te c

usto

m tr

ansp

orts

and

ext

end

LE

IF to

wor

k w

ith

virt

ually

any

mid

dlew

are.

• A

dd e

mer

ging

Web

ser

vice

s st

anda

rds

wit

hout

aff

ectin

g th

e bu

sine

ss lo

gic

of th

e se

rvic

e.

• C

hang

e de

ploy

men

t beh

avio

r, in

sert

tran

spor

ts, o

r m

odif

y m

essa

ge h

andl

ers

wit

hout

cha

ngin

g th

e se

rvic

e co

de.

Dom

ain

LE

IF is

bas

ed o

n in

dust

ry-s

tand

ard

netw

orki

ng, X

ML

, and

Web

ser

vice

s te

chno

logi

es. I

n pa

rtic

u-la

r, L

EIF

pro

vide

s st

rong

sup

port

for

XM

L a

s a

med

ium

of

data

exc

hang

e, e

nabl

ing

inte

rpro

cess

co

mm

unic

atio

n th

at s

pans

pla

tfor

ms

and

lang

uage

s. L

EIF

fit

s w

ell i

nto

the

serv

ices

-ori

ente

d ar

-ch

itect

ure

(SO

A)

appr

oach

to a

pplic

atio

n de

velo

pmen

t.

Too

l Int

egri

ty a

nd S

tand

ards

(as

per

the

Use

r G

uide

Inf

orm

atio

n)

LE

IF p

rodu

cts

conf

orm

to th

e fo

llow

ing

stan

dard

s:

• T

he L

EIF

Web

ser

vice

cod

e ge

nera

tor

uses

WSD

L 1

.1 .

• L

EIF

pro

vide

s do

cum

ents

to c

reat

e co

mpo

nent

s. T

he c

ompo

nent

s pr

oduc

e SO

AP

that

com

-pl

ies

with

the

SOA

P 1.

1 sp

ecif

icat

ion.

• T

he L

EIF

XM

L d

ata

bind

ing

supp

orts

the

mos

t com

mon

and

use

ful f

eatu

res

of th

e M

ay 2

001

CM

U/S

EI-

2004

-TR

-023

57

XM

L S

chem

a re

com

men

dati

on.

• T

he L

EIF

Web

ser

vice

s co

ntai

ner

clos

ely

follo

ws

the

Java

Ser

vlet

spe

cifi

cati

on. H

owev

er, t

he

spec

ific

atio

n re

quir

es th

at a

n im

plem

enta

tion

be w

ritte

n in

the

Java

lang

uage

. The

refo

re, a

C

++

impl

emen

tatio

n ca

nnot

str

ictly

con

form

to a

ll fe

atur

es o

f th

e sp

ecif

icat

ion.

The

Bob

cat

Ref

eren

ce G

uide

con

tain

s in

form

atio

n on

dif

fere

nces

in in

divi

dual

cla

sses

.

• L

EIF

par

ticip

ates

in th

e SO

AP

Bui

lder

s in

itiat

ive

aim

ed a

t gua

rant

eein

g in

tero

pera

bilit

y am

ong

SOA

P im

plem

enta

tions

. To

view

the

Rog

ue W

ave

test

WSD

Ls,

vis

it w

ww

.whi

tem

esa.

com

/inte

rop.

htm

. In

addi

tion,

LE

IF is

ext

ensi

vely

test

ed a

gain

st th

e m

ost

freq

uent

ly u

sed

Java

and

Win

dow

s-ba

sed

(.N

ET

) im

plem

enta

tion

s of

Web

ser

vice

s.

Skill

s B

ase

and

Use

r C

apab

ility

The

use

r ca

pabi

lity

and

skill

s ba

se r

equi

res

the

user

to b

e aw

are

of th

e W

eb s

ervi

ce a

ppli

catio

ns

and

serv

ice-

orie

nted

arc

hite

ctur

e an

d al

so b

e ca

pabl

e of

cre

atin

g th

e W

SDL

for

the

appl

icat

ion

who

se s

ervi

ce h

as to

be

crea

ted.

Con

figu

rati

on C

ontr

ol

The

con

figu

ratio

n an

d th

e de

velo

pmen

t env

iron

men

t set

up

are

wel

l exp

lain

ed in

the

user

gui

de o

f th

e L

EIF

. It s

pans

acr

oss

thre

e st

eps:

• Se

t RW

_HO

ME

(U

NIX

/Lin

ux o

nly)

.

• R

un le

ifva

rs.

• E

nsur

e th

e pr

esen

ce o

f a

JRE

.

Too

l Mod

ular

ity

The

tool

is n

ot h

ighl

y m

odul

ariz

ed; i

t dep

icts

a th

ree-

tier i

nter

face

for t

he S

OA

to b

e de

velo

ped.

Use

r In

terf

ace

A s

impl

e an

d ch

aste

use

r in

terf

ace

of L

EIF

mak

es th

e w

ork

of th

e to

ol e

asily

und

erst

anda

ble

and

also

red

uces

the

com

plex

ity in

volv

ed in

the

user

inte

ract

ion.

58

CM

U/S

EI-

2004

-TR

-023

Res

pons

e T

ime

Fast

est a

vaila

ble

proc

essi

ng o

f W

eb s

ervi

ce m

essa

ges

- up

to 3

00%

net

per

form

ance

gai

n ov

er th

e pr

evio

us r

elea

se L

EIF

1.2

!

Use

r In

terf

ace

Doc

umen

tati

on

LE

IF o

ffer

s a

stan

dard

use

r in

terf

ace

docu

men

tatio

n.

http

://w

ww

.rog

uew

ave.

com

/pro

duct

s/le

if/g

ui.c

fm

Em

plac

emen

t In

tegr

atio

n w

ith

Oth

er T

ools

7

Inte

grat

ion

wit

h O

ther

Too

ls

LE

IF in

clud

es a

ver

sion

of

the

Apa

che

HT

TP

Serv

er o

r M

icro

soft

IIS

pre

conf

igur

ed f

or u

se w

ith

LE

IF in

a d

evel

opm

ent e

nvir

onm

ent.

LE

IF is

cer

tifie

d to

use

Xal

an J

ava

2.4.

1 an

d X

erce

s-J

2.3.

0 fo

r al

l sup

port

ed v

ersi

ons

of th

e JD

K.

Inst

allin

g L

EIF

req

uire

s su

ppor

t of

the

foll

owin

g so

ftw

are,

bas

ed o

n th

e op

erat

ing

syst

em:

• V

isua

l Stu

dio

.NE

T 2

003

• Su

n O

NE

Stu

dio

8

• gc

c 3.

2.3

CMU/SEI-2004-TR-023 59

Appendix C Model Problem and Analysis

The model problem chosen for this project was described as follows:

A major Wall Street investment bank sets out to build a bond pricing system in an effort to stream-line the workflow of its bond trading desk. Currently, bond traders have to send prices for a large number of bonds to several different trading venues, each with its own user interface.

The system that solves the above problem must also minimize the minutiae of pricing all the bonds and provide advanced analytic functionality specific to the bond markets. These capabilities must be provided through a single encapsulated interface.

Classification Scheme Approach

Step 1: Read the problem statement and identify functional and non-functional requirements.

The following requirements can be inferred from the above problem statement:

1. high user interaction

2. integration with the legacy system

3. communication and data exchange mechanism for component2 interaction

4. communication between the C++ and the JAVA applications

Step 2: Map the requirements identified to the integration mechanism, which forms the classifi-cation parameters to be identified in the rapid integration tools.

Analysis of the Functional Requirements

For each of the requirements a specific integration mechanism is suggested as a solution. The mecha-nism will be specific to the particular application. Therefore, the integration mechanisms specified below cannot be generalized for all applications.

2 Here we mean the three components specified by the application: Market Data, Analytics Configuration

and Contribution Server [legacy servers].

60 CMU/SEI-2004-TR-023

Requirements Solutions Integration Mechanism Required

High User Interac-tion

Traders need a very responsive application.

Client application as a Java thick client because of Java’s platform independence and its ability to quickly respond to user input and market data

None

Integration with the legacy system

On the server side, it will inherit leg-acy C++ compo-nents that the sys-tem will utilize. Also, the market data components communicate with the TIBCO3 In-formation Bus (TIB) messaging infrastructure

The following components are to be integrated:

• Market Data Price Feed Server: publishes incom-ing market data to the TIB

• Analytics Engine: performs analytics on incoming market data and broadcasts the modified market data to the TIB

• Contribution Server: performs all communication with trading venues. The trading venues are third- party components not controlled by the bank.

Figure 8: Legacy Market Data Subsystem

JAVA to C++ Translator (Java thick Client talking with C++ Legacy Servers)

3 TIBCO means standard industry-specific messaging infrastructure component.

CMU/SEI-2004-TR-023 61

Figure 9: Legacy Contribution Subsystem

62 CMU/SEI-2004-TR-023

Communication and data exchange mechanism for component4 inter-action

Communication and Data exchange mechanism between sub-components (Thick Client, Market Data and Contribution)

Two gateways to communicate with the legacy servers:

• Pricing Gateway for market data

• Contribution Gateway for sending prices to trading vendors

For instance: With Messaging, we can define separate channels for the different types of pricing data. Then, when a Gateway gets a new piece of data, it will add a message containing that data to the Publish-Subscribe Channel for that data type. Meanwhile, all clients inter-ested in a certain type of data will listen on the channel for that type. In this way, the Gateways can easily send out new data to whoever is interested, without needing to know how many listener applications there are or what they are.

Single point of access through Gateways Mes-saging, Publish and Sub-scribe Channel, JMS (as components are written in JAVA)

Communication between the C++ and the JAVA ap-plication

How to connect the JMS with the standalone C++

Cross language (C++ and JAVA) Messaging Bridge using a combination of Channel Adapters and CORBA.

Messaging Bridge, Chan-nel Adapters and Commu-nication Vehicle between Adapters

4 Here we mean the three components specified by the application, that is, Market Data, Analytics Configu-

ration and Contribution server [Legacy Servers].

CMU/SEI-2004-TR-023 63

Contribution server and the TIBCO based Market Data and Analytics Engine servers?

Analysis of the Non-Functional Requirements Non-functional Requirement

Description

Performance

Here it refers to the scalabil-ity, which can be measured as the number of users it can scale to without noticeable decrease in response time.

One Channel per trader per Bond: Create one Message Channel per-trader per-bond solely for the modified market data of that bond. For example, the market data for bond ABC would be published on channel “Bond ABC” while the modified market data for trader A would be pub-lished on Message Channel “Trader A, Bond ABC,” modified market data for trader B on “Trader B, Bond ABC,” and so on.

64 CMU/SEI-2004-TR-023

Cost

Custom development effort for integration

Effort: 47 person-months for developing the integration components (Refer to Appendix C: Trading Bond COCOTS Estimation Details.)

Hardware/Software Requirements

Below are the hardware and software requirements regarding compo-nents.

1. Analytic Engine and Contribution Server

a. a high-end server class machine with minimum of 512 Mb of RAM

b. Windows 2000 server

2. Traders Desktop Machine (Client):

a. Windows NT, Solaris b. 128 MB of RAM c. Java Virtual Machine

3. TIBCO Information BUS Messaging infrastructure

4. Market Data Price Feed Server

Impacts/Change Analysis on Architecture

The high-level architecture of the system is represented in Figure 10.

CMU/SEI-2004-TR-023 65

Figure 10: Logical View of the System

1. TIBCO Information Bus Messaging infrastructure has been selected to achieve three-way com-munications between Market Data Feed Server, Analytics Engine and Pricing Gateway as shown in Figure 10.

2. Two Java Gateways are used to provide communication between the Market Data Feed Server and the Trading Venues:

a. Pricing Gateway for Market Data Feed Server b. Contribution Server for sending prices to Trader Venues

3. Message Bridge is used to provide communication between JMS used to provide communication between Pricing and Contribution Gateways and TIB (TIBCO Information Bus). This message bridge has C++ and Java Adapters and these adapters communicate with each other through CORBA.

Constraints and Assumptions Made about the Components 1. The system inherits C++ legacy components namely: Market Data Feed Server and Contribution

Server.

2. The system also uses TIBCO Information Bus Messaging Infrastructure as a third-party component.

3. The traders’ venue desktop can run on Windows NT or Solaris Operating System.

66 CMU/SEI-2004-TR-023

Step 3: List the integration mechanisms which are the classification parameters and categorize them into primitive classification types. Categorization of Classification Parameters

Integration Patterns (Primitive Classification Type Parameters)

Primitive Classification Type

Legacy Translator (Java thick Client talking with C++ Legacy Servers)

Legacy Integration

Gateways Application Integration

Messaging (JMS) Middleware Integration

Publish and Subscribe Channel Middleware Integration

Messaging Bridge Application Integration

Channel Bridge Application Integration

Communication Vehicle between Adapters Application Integration

From this table we can infer that the current scenario is a composite of three primitive classification types, namely

• Legacy Integration

• Application Integration

• Middleware Integration

Step 4: Identify the rapid integration tools needed to quickly solve this problem.

In this step we try to represent the scenario as a set of classification parameters. Here we have the integration mechanisms that serve as the classification parameters.

Mathematically, scenario can be expressed as

Scenario1 = {Legacy Translator, Gateways, Messaging (JMS), Publish & Subscribe Channel, Mes-saging Bridge, Channel Bridge, Communication Vehicle between Adapters}

The parameters identified using this scenario form the elements of the primitive classification type.

1. Legacy Integration = {Legacy Translator}

2. Application Integration = {Gateways, Messaging Bridge, Channel Bridge, Communication Ve-hicle between Adapters}

3. Middleware Integration = {Messaging (JMS), Publish and Subscribe Channel}

CMU/SEI-2004-TR-023 67

For the current scenario the parameters assume following values:

1. Legacy Translator=“Java to C++ Translator”

2. Gateways = “Java Gateways”

3. Messaging Bridge=“Bridge C++ Messaging system to JAVA Messaging System”

4. Channel Bridge=“C++ TIB Adapter & JMS Adapter”

5. Communication Vehicle between Adapters =“CORBA”

6. Messaging (JMS) =“IBM MQ Series”

7. Publish and Subscribe Channel=“Channel for different types of pricing data with Gateways as Publisher and Clients as Subscriber”

From the analysis done to classify the rapid integration tool we determine it to be a set of the combi-nations of the primitive classification types:

RIT for Scenario = {Legacy, Application, Middleware}

Similarly when we generalize it

RIT for Scenarion = {Primitive Classification Type *}

Step 5: Select tools.

Through use of the Tool Classification Matrix the following tools are identified to support this inte-gration.

1. Microsoft BizTalk Server

2. IBM WebSphere

3. Pervasive Data Junction

68 CMU/SEI-2004-TR-023

Tools Classification Matrix

Name of the Tool Classification based on Primitive Integration Types

Leg

acy

Inte

grat

ion

Mid

dlew

are

Inte

grat

ion

Serv

ice-

Ori

ente

d In

tegr

a-ti

on

App

licat

ion

Inte

grat

ion

Web

-Bas

ed I

nteg

rati

on

Pervasive Data Junction

RogueWave's LEIF

IBM Rational Rapid Developer

Microsoft SQL Server

Host Integration Server

Microsoft BizTalk Server

IBM WebSphere Business Integration

Artix Relay

Artix Encompass

Artix Mainframe

PiiE Smart Client

PiiE Fusion Server

InterSystem Ensemble

Jboss

CMU/SEI-2004-TR-023 69

Appendix D Commercial Off-the-Shelf Components

This section describes the Customized Dukascopy Data Feed Components (CDDF). http://www.dukascopy.com/english/ddf_main/rdata/

Java Feeder Component: The Feeder components are commercial off-the-shelf (COTS) products provided by Dukascopy. These components connect themselves to Dukascopy Market Machine data source and supply data every 10 seconds to the software application connected to it. Dukascopy Market Machine data source supplies data on liquid trading instruments.

The data has the following format: stockId – integer Value – double tickVolume – integer (on every instrument) where stockID is the ID of trading instrument set by the user Value is an average 10 sec price value.

Besides providing real-time data, this component can also transfer historical data going back three days (nearly 22000 of 10 sec ticks) that can be used to fill in occasional gaps in the database.

Component Specification:

The Feeder component provides interfaces and methods listed below. These can be used by the appli-cation program to capture the data collected by this component from Dukascopy Market Machine data source:

1. DataListener Interface

onNewTick(int id, double value, int volume): This method provides the data that is fetched from the Dukascopy Market Machine data source.

2. TickerListener Interface

a. onNewTick(int id, double value, int volume): This method provides the data that is fetched from the Dukascopy Market Machine data source.

b. onNewConnection(Connector conn)

3. addQuote(int id, String code) method in FeederConnector Class: This method allows the applica-tion program to add a specific trading instrument for which the data has to be collected.

4. removeQuote(String code) method in FeederConnector Class: This method allows the applica-tion program to remove a specific trading instrument for which the data has to be collected.

70 CMU/SEI-2004-TR-023

5. setDataListener(DataListener dl)) method in FeederConnector Class: This method provides the data that is fetched from the Dukascopy Market Machine data source. It eventually uses the on-NewTick(int id, double value, int volume) method to get the data.

6. connect() method in FeederConnector: This method initiates the connection of this component to the Dukascopy Market Machine data source.

Figure 11 below shows the interfaces and methods within those interfaces which are accessible to external programs.

Figure 11: Feeder Component Specification

Component Realization:

The feeder component is implemented using the following Java classes and interfaces.

1. FeederConnector

2. ConnectorWorker

3. Connector: Protocol realization

4. DataListener: Client interface for working with data

5. TickerListener

6. TickerWorker

Connector

DataListener

onNewTick()

TickerWorker

setListener()onNewConnection()onNewData() onNewCommand()

TickerListener

onNewTick() onNewConnection()

FeederConnector

addQuote(int id, String code) removeQuote(String code) setDataListener(DataListener dl) connect()

CMU/SEI-2004-TR-023 71

VC++ MarketDataFeed Component:

The following are the VC++ files that define the responsibility of the VC++ Component:

1. CConn.vcproj

This is the main project file for VC++ projects generated using an Application Wizard. It con-tains information about the version of Visual C++ that generated the file and information about the platforms, configurations, and project features selected with the Application Wizard.

2. CConn.idl

This file contains the IDL definitions of the type library, the interfaces and co-classes defined in the project. This file will be processed by the MIDL compiler to generate

C++ interface definitions and GUID declarations (CConn.h)

3. CCoCConn.vcproj

This is the main project file for VC++ projects generated using an Application Wizard. It con-tains information about the version of Visual C++ that generated the file, and information about the platforms, configurations, and project features selected with the Application Wizard.

4. CConn.idl

This file contains the IDL definitions of the type library, the interfaces and co-classes defined in the project.

DataListener DataListener

Connector Connector

Worker

TickerWorker

TickerListener TickerListener

72 CMU/SEI-2004-TR-023

This file will be processed by the MIDL compiler to generate the following:

C++ interface definitions and GUID declarations

(CConn.h)

GUID definitions (CConn_i.c)

A type library (CConn.tlb)

Marshaling code (CConn_p.c and dlldata.c)

5. CConn.h

This file contains the C++ interface definitions and GUID declarations of the items defined in CConn.idl. It will be regenerated by MIDL during compilation.

6. CConn.cpp

This file contains the object map and the implementation of your DLL's exports.

7. CConn.rc

This is a listing of all of the Microsoft Windows resources that the program uses.

8. CConn.def

This module-definition file provides the linker with information about the exports required by the DLL. It contains exports for

DllGetClassObject DllCanUnloadNow GetProxyDllInfo DllRegisterServer DllUnregisterServer

Other standard files

9. StdAfx.h, StdAfx.cpp

These files are used to build a precompiled header (PCH) file named CConn.pch and a precom-piled types file named StdAfx.obj.

10. Resource.h

This is the standard header file that defines resource IDs: Proxy/stub DLL project and module definition file.

11. CConnps.vcproj

This file is the project file for building a proxy/stub DLL if necessary.

The IDL file in the main project must contain at least one interface and you must first compile the IDL file before building the proxy/stub DLL. This process generates dlldata.c, CConn_i.c and CConn_p.c , which are required to build the proxy/stub DLL.

CMU/SEI-2004-TR-023 73

12. CConnps.def

This module definition file provides the linker with information about the exports required by the proxy/stub.

Other notes:

The MFC Support option builds the Microsoft Foundation Class libraries into your skeleton applica-tion, making MFC classes, objects and functions available to you.

Issues 1. If the client process is not killed properly, the Java component will still deliver the data to the

client application. This state prevents the client application from re-establishing the lost connec-tion to properly terminate the data stream.

74 CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 75

Appendix E Trading Bond System COCOTS Estimation Details

The approach followed is strictly based on the COCOTS estimation model proposed by Christopher M. Abts and Barry W. Boehm [Abst 00]. The standard COCOTS calibration tables are used for the calibrated parameter values for each cost driver in the model. The corresponding parameter value for each driver is fed into the spreadsheet tool—COCOTS calculator.

Assumptions 1. The values of very high, low, and so forth, have been determined based on a heuristic approach

rather than on previous data collection.

2. The KSLOC is assumed to be based on the programming experience of the team with the prior knowledge of the domain addressed here. The SLOC for developing a glue code for integrating the C++ and Java Components using JNI is found to be approximately 1000 SLOC [1 KSLOC].

The component that is the glue code for the integrating C++ and Java is assumed to be devel-oped using JNI. We realize that the excerpts taken from the article on Junc++ion (http://www.codemesh.com/en/CodemeshWhitepaper.pdf ) demonstrate that JNI requires a huge number of lines of code.

“If the programmer were trying to write an application to display a Java Swing dialog box from C++ and store the user’s input in C++ using JNI to communicate between C++ and Java, about 200 lines of JNI code would be required.”

3. Since there are no real-world customers, there is a very minimal requirement change for this in-tegration scenario and hence the BRAK % is assumed to be 1.

4. The Normal Labor Cost here refers to the Software Engineers in any company that will be in-volved in the development.

Constraints

Currently, we have one option for C++ and Java components. Also, the Trading Bond System here addressed is restricted to the legacy integration of components

Cost Drivers Selection

The following table presents the values selected and the reasons for their selection for the various cost drivers of the COCOTS estimation model.

76 CMU/SEI-2004-TR-023

Category Cost Drivers Value Why?

Integration Personal Drivers

1 COTS Integrator Ex-perience with Product

VL Two months of experience with the products

2 COTS Integrator Per-sonnel Capability

L Two months of experience with the domain

3 Integrator Experience with COTS Integration Processes

L Organizational level [Professional Development Cen-ter] process for COTS integration is not defined.

4 Integrator Personnel Continuity

N There will be a rotation of people every year in the Professional Development Center, as it is an educa-tional environment.

COTS Component Drivers

1 COTS Product Maturity H The product has high time on market.

2 COTS Supplier Product Extension Willingness

L The products we consider here are standard C++ and Java components available on the net; hence the num-ber and nature of changes are very minimal.

3 COTS Product Interface Complexity

N Since most of the APIs of the components are well de-fined, consistently applied, and clear, they can easily be used to interface with the glue code.

4 COTS Supplier Product Support

H The level of available support is high; a detailed ex-planation of the components to be integrated is available.

5 COTS Supplier Pro-vided Training and Documentation

N Nominal documentation is provided for the scenario considered here.

6 COTS Product Volatility L Only one release is expected.

CMU/SEI-2004-TR-023 77

Application/System Drivers

1 Constraints on System/Subsystem Reliability

N This is not a mission-critical system; there are backup servers to recover the lost data.

2 Application Interface Complexity.

L Use of standard communication mechanisms such as APIs reduces the application interface complexity.

3 Constraints on System/Subsystem Technical Performance

VH The analytic engine handles the real-time market data and feeds it to trader’s desktop.

4 System Portability VH The traders’ desktops might be running on different operating systems.

Nonlinear Scale Factor

1 Application Architec-tural Engineering

L Simple paper Analysis of the architecture of the system will be done for the currently selected scenario.

78

CM

U/S

EI-

2004

-TR

-023

CO

TS

Inte

grat

or C

alcu

lato

r

CM

U-W

est

C

OT

S In

tegr

atio

n C

ost

Cal

cula

tor

V2.

0

(C

OC

OT

S G

lue

Cod

e Su

b-m

odel

)

Cal

i-br

atio

n T

able

s:

Dri

ver

R

atin

g 1

2 3

4 5

6 7

8 9

10

11

12

13

14

AC

IEP

A

CIP

C

AX

ICP

A

PC

ON

A

CP

MT

A

CSE

W

AP

CP

X

AC

PP

S A

CP

TD

A

PV

OL

A

CR

EL

A

AC

PX

A

CP

ER

A

SPR

T

V

L

1.34

1.

60

1.

58

1.45

1.20

0.

71

L

1.

16

1.27

1.

12

1.26

1.

20

1.07

0.

82

1.14

1.

09

0.84

0.

88

0.84

N

1.

00

1.00

1.

00

1.00

1.

00

1.00

1.

00

1.00

1.

00

1.00

1.

00

1.00

1.

00

1.00

H

0.

86

0.79

0.

89

0.80

0.

83

0.94

1.

22

0.88

0.

91

1.19

1.

14

1.19

1.

11

1.07

V

H

0.75

0.

62

0.79

0.

63

0.69

0.

88

1.48

0.

77

0.84

1.

33

1.30

1.

42

1.22

1.

14

L

inea

r Sc

alin

g F

acto

r

Non

linea

r Sc

alin

g F

acto

r

A

AA

RE

N

12

.0

VL

L

N

H

V

H

CM

U/S

EI-

2004

-TR

-023

79

4.

00

3.00

2.

00

1.00

0.

00

Inpu

ts:

Dri

ver

1 2

3 4

5 6

7 8

9 10

11

12

13

14

AC

IEP

A

CIP

C

AX

IC P

AP

CO

N

AC

PM

T

AC

SEW

A

PC

PX

A

CP

PS

AC

PT

D

AP

VO

L

AC

RE

L

AA

CP

X

AC

PE

R

ASP

RT

V

alue

1.

34

1.27

1.

12

1.00

0.

83

1.07

1.

00

0.88

1.

00

0.84

1.

00

0.84

1.

22

1.14

Sc

alin

g F

acto

r

Si

zing

Lab

or C

ost

A

AR

EN

K

SLO

C

BR

AK

(%)

R

ate

($/P

r-m

t)

Val

ue

3.

00

1.

00

1

7000

.00

Out

puts

:

E

SIZ

E

B

EM

1-14

Est

imat

ed E

ffor

t (P

rs-m

ts)

Est

imat

ed C

ost

($)

1.

01

1.00

1.

46

17

.63

$1

23,4

07

80

CM

U/S

EI-

2004

-TR

-023

CMU/SEI-2004-TR-023 81

Appendix F Project Details

This appendix contains the detailed work breakdown structure (WBS) and project details.

Estimated Effort hours: 3x10x24 = 720 person hours

No. of Team members = three No. of hours per week per team member = 10 hours No. of months = six (equivalent to 24 weeks)

Project Planned Start Date: Fri 1/23/04 Project Planned Finish Date: Wed 6/23/04

The overall schedule of the planned project is given in detail in WBS. Here we illustrate with a sim-ple timeline the overall schedule of the project.

Planned Schedule Project Start Project End

▼ ▼Jan 04 Feb 04 Mar 04 Apr 04 May 04 Jun 04

▲ Milestone 1

▲ Milestone 2

Actual Schedule Project Start Project End

▼ ▼Jan 04 Feb 04 Mar 04 Apr 04 May 04 Jun 04 Jul 04

▲ Milestone 1

▲ Milestone 2

• Project Start denotes the actual project start date of the rapid integration tools project.

• Milestone 1 implies the completion of Task1 - This included identification of tools and coming up with a classification scheme for them.

• Milestone 2 denotes the completion of Task2 - This includes identifying the model problem and getting hands on experience in evaluating the tools which would help in solving the model problem.

82 CMU/SEI-2004-TR-023

• Project End denotes the submission of the evaluation results in the form of a technical report and expressing the process of evaluation as a framework that can be extended to any model problem.

The line in the actual schedule denotes where we were when we were writing this technical report.

Similarly, the estimated effort into the project also increased from 720 person hours to 840 person hours.

Estimation:

The above WBS is based on the rapid integration tools document provided by the SEI before the start of the project. The project is divided into three tasks which have deliverables associated with each of them. The milestones are based completely on the three tasks. Each Task was allocated two months out of the total six months for the project.

Actual Progress:

However, as shown in the actual progress timeline, Task 1 took almost three months for completion, Task 2 took another three months to complete, and Task 3 is currently underway at the time of writing this report.

The primary reasons for schedule slippage are multiple commitments of team members on other pro-jects, and the fewer number of hours allocated for the elective.

Table 8: Milestones and Schedule of the Project

Milestones Expected Date Revised Date of Submission

Actual Date of Submission

Task 1 - Survey and clas-sify the tools-

• List of Rapid Integra-tion Tools

2/5/2004 6/7/2004 5/18/2004

• Classification Scheme 2/17/2004 6/09/2004 6/5/2004

• Classified List of Tools

3/02/2004 6/10/2004 6/5/2004

Task 2 - Evaluate the tools using a model problem

• Preliminary Evalua-tion Scheme

3/19/2004 6/16/2004

• Model Problem Defi-nition

4/26/2004 6/16/2004 5/18/2004

CMU/SEI-2004-TR-023 83

• Reports detailing evaluation of tools’ applicability to the model problem

5/11/2004 6/24/2004 5/18/2004

Task 3 - Develop and document general evalua-tion criteria

• Documented evalua-tion criteria for rapid integration tools

6/1/2004 5/28/2004

84 CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 85

Glossary of Technical Terms

Term Description

EAI “Acronym for enterprise application integration. EAI is the unrestricted sharing of data and business processes throughout the networked applications or data sources in an or-ganization. Early software programs in areas such as inventory control, human re-sources, sales automation and database management were designed to run independ-ently, with no interaction between the systems. They were custom built in the technology of the day for a specific need being addressed and were often proprietary systems. As enterprises grow and recognize the need for their information and applica-tions to have the ability to be transferred across and shared between systems, companies are investing in EAI in order to streamline processes and keep all the elements of the enterprise interconnected.

There are four major categories of EAI:

1. Database linking: databases share information and duplicate information as needed.

2. Application linking: the enterprise shares business processes and data between two or more applications.

3. Data warehousing: data is extracted from a variety of data sources and channeled into a specific database for analysis.

4. Common virtual system: the pinnacle of EAI; all aspects of enterprise computing are tied together so that they appear as a unified application.”

http://www.webopedia.com/TERM/E/EAI.html

B2Bi Business-to-Business Integration

Legacy Sys-tem

“A computer system or application program which continues to be used because of the cost of replacing or redesigning it and often despite its poor competitiveness and com-patibility with modern equivalents. The implication is that the system is large, mono-lithic and difficult to modify”

http://computing-dictionary.thefreedictionary.com/Legacy%20system

Adapters “Adapters and Connectors are pieces of software that are used in the integration of com-ponent-based applications and serve as a “wrapper“ that mediates access to an applica-tion that was not developed with integration in mind, including legacy applications”

86 CMU/SEI-2004-TR-023

http://eai.ittoolbox.com/nav/t.asp?t=346&p=347&h1=346&h2=347

Service-Oriented Inte-gration

“Service-Oriented Integration (SOI) leverages open standards, loose coupling, and dy-namic description and discovery capabilities of Web Services to reduce the complexity, cost, and risk of integration.”

http://www.zapthink.com/cluster.html?id=soi

Web Services “Web Services refers to the technologies that allow for making connections. Services are what you connect together using Web Services. A service is the endpoint of a connec-tion. Also, a service has some type of underlying computer system that supports the connection offered. The combination of services—internal and external to an organiza-tion—make up a service-oriented architecture.”

http://www.service-architecture.com/web-services/articles/web_services_definition.html

ALE Stands for Application Embedding and Linking. “ALE allows behaviors between com-ponents and applications to be linked on a single-screen. Users are able to drill within applications, as well as from one application to another, without changing focus.

ALE overcomes the limitations of HTML-based Web applications where any embedded link typically brings up a new page with no contextual link between the various Web pages.”

http://www.dharbor.com/products/psc_feat.html

JMX “Java Management Extensions (JMX) technology provides the tools for building dis-tributed, Web-based, modular and dynamic solutions for managing and monitoring de-vices, applications, and service-driven networks. By design, this standard is suitable for adapting legacy systems, implementing new management and monitoring solutions, and plugging into those of the future “

http://java.sun.com/products/JavaManagement/

CICS “Short for Customer Information Control System, a TP monitor from IBM that was originally developed to provide transaction processing for IBM mainframes. It controls the interaction between applications and users and lets programmers develop screen dis-plays without detailed knowledge of the terminals being used.”

http://www.webopedia.com/TERM/C/CICS.html

SOAP “Short for Simple Object Access Protocol, a lightweight XML-based messaging proto-col used to encode the information in Web service request and response messages before

CMU/SEI-2004-TR-023 87

sending them over a network.

SOAP messages are independent of any operating system or protocol and may be trans-ported using a variety of Internet protocols, including SMTP, MIME, and HTTP.”

http://www.webopedia.com/TERM/S/SOAP.html

IIOP “Short for Internet Inter-ORB Protocol, a protocol developed by the Object Management Group (OMG) to implement CORBA solutions over the World Wide Web. IIOP enables browsers and servers to exchange integers, arrays, and more complex objects, unlike HTTP, which only supports transmission of text.”

http://www.webopedia.com/TERM/I/IIOP.html

WSDL “Short for Web Services Description Language, an XML-formatted language used to describe a Web service's capabilities as collections of communication endpoints capable of exchanging messages. WSDL is an integral part of UDDI, an XML-based worldwide business registry. WSDL is the language that UDDI uses. WSDL was developed jointly by Microsoft and IBM.”

http://www.webopedia.com/TERM/W/WSDL.html

LDAP “Short for Lightweight Directory Access Protocol, a set of protocols for accessing in-formation directories. LDAP is based on the standards contained within the X.500 stan-dard, but is significantly simpler. And unlike X.500, LDAP supports TCP/IP, which is necessary for any type of Internet access. Because it's a simpler version of X.500, LDAP is sometimes called X.500-lite.”

http://www.webopedia.com/TERM/L/LDAP.html

End-to-end Encryption

“The encryption of information at its origin and decryption at its intended destination without any intermediate decryption.”

http://www.its.bldrdoc.gov/fs-1037/dir-014/_2016.htm

XML “Short for Extensible Markup Language, a specification developed by the W3C. XML is a pared-down version of SGML, designed especially for Web documents. It allows de-signers to create their own customized tags, enabling the definition, transmission, vali-dation, and interpretation of data between applications and between organizations.”

http://www.webopedia.com/TERM/X/XML.html

Microsoft CLR

“The Microsoft CLR Debugger is intended as an interim tool for debugging applications written and compiled for the common language runtime.”

88 CMU/SEI-2004-TR-023

http://msdn.microsoft.com/library/default.asp?url=/library /en-us/cptutorials/html/the_net_sdk_debugger.asp

XSLT “Short for Extensible Style Language Transformation, the language used in XSL style sheets to transform XML documents into other XML documents. An XSL processor reads the XML document and follows the instructions in the XSL style sheet, then it outputs a new XML document or XML-document fragment. This is extremely useful in e-commerce, where the same data need to be converted into different representations of XML. Not all companies use the exact same programs, applications and computer sys-tems.”

http://www.webopedia.com/TERM/X/XSLT.html

SMTP “Short for Simple Mail Transfer Protocol, a protocol for sending e-mail messages be-tween servers. Most e-mail systems that send mail over the Internet use SMTP to send messages from one server to another; the messages can then be retrieved with an e-mail client using either POP or IMAP. In addition, SMTP is generally used to send messages from a mail client to a mail server. This is why you need to specify both the POP or IMAP server and the SMTP server when you configure your e-mail application.”

http://www.webopedia.com/TERM/S/SMTP.html

HTTP “Short for HyperText Transfer Protocol, the underlying protocol used by the World Wide Web. HTTP defines how messages are formatted and transmitted, and what actions Web servers and browsers should take in response to various commands. For example, when you enter a URL in your browser, this actually sends an HTTP command to the Web server directing it to fetch and transmit the requested Web page.”

http://www.webopedia.com/TERM/H/HTTP.html

PKI “Short for public key infrastructure, a system of digital certificates, Certificate Authori-ties, and other registration authorities that verify and authenticate the validity of each party involved in an Internet transaction. PKIs are currently evolving and there is no single PKI nor even a single agreed-upon standard for setting up a PKI. However, nearly everyone agrees that reliable PKIs are necessary before electronic commerce can be-come widespread.”

http://www.webopedia.com/TERM/P/PKI.html

J2EE “Short for Java 2 Platform Enterprise Edition. J2EE is a platform-independent, Java-centric environment from Sun for developing, building and deploying Web-based enter-prise applications online. The J2EE platform consists of a set of services, APIs, and pro-tocols that provide the functionality for developing multi-tiered, Web-based applica-

CMU/SEI-2004-TR-023 89

tions.”

http://www.webopedia.com/TERM/J/J2EE.html

JSP “Short for Java Server Page. A server-side technology, Java Server Pages are an exten-sion to the Java servlet technology that was developed by Sun. JSPs have dynamic scripting capability that works in tandem with HTML code, separating the page logic from the static elements -- the actual design and display of the page—to help make the HTML more functional (i.e., dynamic database queries).”

http://www.webopedia.com/TERM/J/JSP.html

JCA “The J2EE Connector architecture provides a Java technology solution to the problem of connectivity between the many application servers and today's enterprise information systems (EIS).”

http://java.sun.com/j2ee/connector/overview.html

EJB “Enterprise JavaBeans (EJB) is a Java API developed by Sun Microsystems that defines component architecture for multi-tier client/server systems. EJB systems allow develop-ers to focus on the actual business architecture of the model, rather than worry about endless amounts of programming and coding needed to connect all the working parts. This task is left to EJB server vendors. Developers just design (or purchase) the needed EJB components and arrange them on the server.”

http://www.webopedia.com/TERM/E/Enterprise_JavaBeans.html

JAAS “The Java Authentication and Authorization Service (JAAS) is a set of APIs that enable services to authenticate and enforce access controls upon users. It implements a Java technology version of the standard Pluggable Authentication Module (PAM) framework, and supports user-based authorization.”

http://java.sun.com/products/jaas/

Aspect Ori-ented Pro-gramming

“Aspect-oriented programming (AOP) is a new programming technique that allows pro-grammers to modularize crosscutting concerns (behavior that cuts across the typical di-visions of responsibility, such as logging). AOP introduces aspects, which encapsulate behaviors that affect multiple classes into reusable modules.”

http://www-106.ibm.com/developerworks/java/library/j-aspectj/

90 CMU/SEI-2004-TR-023

CMU/SEI-2004-TR-023 91

References/Bibliography

URLs are valid as of the publication date of this document.

[Abst 00] Abst, C.; Boehm, B.; & Clark, B. COCOTS: A Software Integration Cost Model. Los Angeles, CA: USC Center for Software Engineer-ing 2000. http://sunset.usc.edu/publications/TECHRPTS /2000 /usccse2000-501/usccse2000-501.pdf

[Brownsword 04] Brownsword, L.; Carney, D. J.; Fisher, D.; Lewis, G.; Meyers, C.; Morris, E. J.; Place, R. H.; Smith, J.; & Wrage, L. Current Perspec-tives on Interoperability (CMU/SEI-2004-TR-009, ADA421613). Pittsburgh, PA: Software Engineering Institute, Carnegie Mellon University, 2004. http://www.sei.cmu.edu/publications/documents /04.reports/04tr009.html

[IONA 02] ONA Technologies PLC. IONA’s Orbix E2A Enterprise Integration Technology. Waltham, MA: IONA Technologies PLC, 2002.

[Jaccheri 02] Jaccheri, L. & Torchiano, M. “Classifying COTS Products,” 246-255. Software Quality – ECSQ 2002. Quality Connection – 7th European Conference on Software Quality. Helsinki, Finland, June 9-13, 2002. Proceedings, Lecture Notes in Computer Science, 2349. Berlin, Germany: Springer-Verlag, 2002.

[Johnson 88] Johnson, A. M. & Malek, M. “Survey of Software Tools for Evaluat-ing Reliability, Availability, and Serviceability.” Computing Surveys, 20, 4, (Dec. 1988) 227-269.

[Kasunic 03] Kasunic, Mark. “Measuring Systems Interoperability.” Conference on the Acquisition of Software-Intensive-Systems. Arlington, VA, January 28-30, 2003. Pittsburgh, PA: Software Engineering Insti-tute, Carnegie Mellon University 2003.

[Ncube 02] Ncube, C. & Dean, J. “Limitations of Current Decision-Making Techniques in the Procurement of COTS Software Components,” 176-187. Proceedings of the First International Conference of COTS-Based Software Systems, ICCBSS 2002. Orlando, FL, Febru-ary 4-6, 2002. Berlin, Germany: Springer-Verlag, 2002.

92 CMU/SEI-2004-TR-023

[Powell 97] Powell, A., et al. “Evaluating Tools to Support Component Based Software Engineering, 80-89. Proceedings of the Fifth International Symposium on Assessment of Software Tools and Technologies. Pittsburgh, PA, June 2-5, 1997. Los Alamitos, CA: IEEE Computer Society Press, 1997.

[Sai 04] Vijay, S. COTS Acquisition Evaluation Process: Preacher’s Practice (CMU/SEI-2004-TN-001, ADA421675). Pittsburgh, PA: Software Engi-neering Institute, Carnegie Mellon University, 2004. http://www.sei.cmu.edu/publications/documents /04.reports/04tn001.html

[Schmidt 02] Schmidt, Marty. J. Business Case Essentials: A Guide to Structure and Content. West Sussex, UK: Matrix Solutions Ltd, 2002. http://www.eitoolkit.com/tools/initiation /Business_Case_Essentials.pdf

[Simon 03] Simon, Jonathan. “Case Study: Bond Trading System.” Chapter 13, Hohpe, Gregor & Woolf, Bobby. Enterprise Integration Patterns. Boston, MA: Addison-Wesley, 2004.

[Vu 02] Vu, John. “The eCommerce Capability Model (eCCM): A Framework To Implement eCommerce Successfully.”(tutorial). SEPG 2000: Software En-gineering Process Group Conference, Phoenix, AZ, February 18-21, 2002.

[Wallnau 91] Wallnau, K. & Feiler, P. Tool Integration and Environment Architectures (CMU/SEI-91-TR-011, ADA237810). Pittsburgh, PA: Software Engineer-ing Institute, Carnegie Mellon University, 1991. http://www.sei.cmu.edu/publications/documents /91.reports /91.tr.011.html

[Wallnau 02] Wallnau, K.; Hissam, S.; & Seacord, R. Building Systems from Commercial Components. Boston, MA: Addison Wesley, 2002.

REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704-0188

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503.

1. AGENCY USE ONLY

(Leave Blank)

2. REPORT DATE

December 2004

3. REPORT TYPE AND DATES COVERED

Final 4. TITLE AND SUBTITLE

Rapid Integration for Rapid Application Development

5. FUNDING NUMBERS

F19628-00-C-0003 6. AUTHOR(S)

Amit Midha, Ravindra Singh, Lakshmi Pratha Hari 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

Software Engineering Institute Carnegie Mellon University Pittsburgh, PA 15213

8. PERFORMING ORGANIZATION REPORT NUMBER

CMU/SEI-2004-TR-023

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

HQ ESC/XPK 5 Eglin Street Hanscom AFB, MA 01731-2116

10. SPONSORING/MONITORING AGENCY REPORT NUMBER

ESC-TR-2004-023

11. SUPPLEMENTARY NOTES

12A DISTRIBUTION/AVAILABILITY STATEMENT

Unclassified/Unlimited, DTIC, NTIS

12B DISTRIBUTION CODE

13. ABSTRACT (MAXIMUM 200 WORDS)

This report investigates the rapid integration tools available in the current market. These tools aid in the rapid integration of software systems and components. The research centers on a model problem that requires such a tool to address legacy integration challenges. The report presents a generic evaluation framework for identifying and evaluating rapid integration tools and an evaluation of three identified tools. This evaluation engaged selected evaluation criteria based on the demands of the model problem. A process reference is also included; this forms the guidelines for identification and evaluation of the tools with respect to other model problems.

14. SUBJECT TERMS

rapid integration, integration tools, rapid application development

15. NUMBER OF PAGES

107 16. PRICE CODE

17. SECURITY CLASSIFICATION

OF REPORT

Unclassified

18. SECURITY CLASSIFICATION OF THIS PAGE

Unclassified

19. SECURITY CLASSIFICATION OF ABSTRACT

Unclassified

20. LIMITATION OF ABSTRACT

UL

NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. Z39-18 298-102