Chapter 7. Empowering the Internet-Enabled Information Infrastructure

771 - Chapter 7. Empowering the Internet-Enabled Information Infrastructure <blockquote> <p><script> function OpenWin(url, w, h) { if(!w) w = 400; if(!h) h = 300; window. open (url, "_new", "width=" + w + ",height=" + h + ",menubar=no,toobar=no,scrollbars=yes", true); } function Print() { window.focus(); if(window.print) { window.print(); window.setTimeout('window.close();',5000); } } </script><span></span></p> <table width="100%" cellpadding="0" cellspacing="0" border="0"><tr valign="top"></tr></table> <table width="100%" height="20" border="0" cellpadding="0" cellspacing="1"><tr></tr></table> <table width="100%" border="0" cellspacing="0" cellpadding="0"> <tr valign="top"> <td align="center"><table width="95%"><tr><td align="left"> <table width="100%" border="0" cellspacing="0" cellpadding="2"> <tr><td valign="top" height="5"><img src="/books/2/551/1/html/2/images/pixel.gif" width="1" height="5" alt="" border="0"></td></tr> <tr> <td><b><font size="1" color ="#0000FF">Team-Fly<img border="0" src="/books/2/551/1/html/2/Fly-Logo.gif" width="81" height="25"></font></b></td> <td valign="top" align="right">     </td> </tr> </table> <hr size="1"> <table width="100%" border="0" cellspacing="0" cellpadding="5"> <tr> <td valign="top" width="76" rowspan="4"><img src="/books/2/551/1/html/2/images/0130409510/0130409510_xs.jpg" width="76" height="95" border="0"></td> <td valign="top">Internet-Enabled Business Intelligence<br>By William A. Giovinazzo<br> </td> </tr> <tr><td>Table of Contents</td></tr> <tr><td></td></tr> <tr><td valign="bottom">Part 3.  The Software of the Internet</td></tr> </table> <hr size ="1"> <br><table width="100%" border="0" cellspacing="0" cellpadding ="0"><tr><td valign="top"> <h2>Chapter 7. Empowering the Internet-Enabled Information Infrastructure</h2> <blockquote> <p><span>The galling thing about the generally poor quality of much current software is that there is no extrinsic reason for it; perfection is, in principle, possible.</span></p> <p><span>It is only our own human frailties that limit the quality of software. We humans are not particularly adept at dealing with extremely complex situations. On the surface it may appear that large programs are rather like small onesonly bigger. Nothing, however, could be further from the truth. The complexity of programs grows at least as the square of their size. It doesn't take a very large program to be beyond our human limitations.</span></p> <p><span>William A. Wulf Comments on "Current Practice"</span><br><span>Research Directions in Software Technology</span><sup>[1]</sup></p> <blockquote><p><sup>[1]</sup> Wegner, Peter (Ed.), <span>Research Directions in Software Technology</span>, The Massachusetts Institute of Technology, 1979. Used by permission.</p></blockquote> </blockquote> <p>Software, that's what it's all about. The fastest , most advanced computer hardware in the world won't do you a bit of good without the software to drive it. While this may seem obvious, for a very long time people simply did not get it. Many still don't.</p> <p>In my first programming position, I wrote software to run on a Data General computer. At the time, the company had its own proprietary operating system, a pretty darn good operating system at that. The system was so good that several companies attempted to purchase just the operating system. No big deal today, right? Back then, Data General refused to sell the operating system without the hardware. It was so adamant in its opposition that it fought and won a lawsuit in which it was accused of restraining trade.</p> <p>Several years later I worked for a company that was founded by several Data General engineers . When the company's market began to flounder, I discussed with several of the marketing organizations the idea of attacking the database server market. This was back in the late 1980s when there was no clear leader in that market. At the time, the company was too focused on delivering a system based on a newer chip set to be bothered. After all, I was talking about using the old hardware to go after a market with nothing more than new software.</p> <p>About 2 years ago, Data General was purchased by EMC. The server line, which eventually moved to UNIX, is now dead. My second company has faded into history. The company went belly up in the early 1990s, reduced to an imperceptible blip in the evolution of information technology. Both companies missed the point. Hardware no longer drives our industry. Instead, we are driven by the functionality of the software. The hardware has become little more than a commodity.</p> <p>There is a very old, very dumb joke. A man, after 30 years of marriage , is rummaging around in his wife's closet and finds an odd-looking box. In the box he discovers two ears of corn and $300,000 in cash. He confronts his wife with his discovery. She explains that every time she had been unfaithful to him, she saved an ear of corn. Taking this in stride, he reasons that while not pleased with the discovery in the grand scheme of things, two indiscretions in the course of 30 years is not as terrible as it could be. He presses her, however, to explain the $300,000. "Well dear," she replies, "that's from trading in corn commodities."</p> <p>A commodity is something in which there is little distinction, such as the corn in our story. I know farmers who can discuss at great length and in great detail the differences in corn. For most people, though, corn is corn. The distinctions made by a connoisseur are lost on us. The same is true with hardware. While the IT department may be interested in the difference between one hardware platform and another, to most end users, there is little difference.</p> <p>Simple economics is engendering a process of commoditization in the computer industry. Examine the application technology stack presented in Figure 7.1. As each layer in the stack becomes standardized, the layers beneath become more of a commodity. It began with the hardware, or perhaps more accurately, with the operating system layer. At one time, hardware was the main consideration in choosing an application. The selection of one vendor or another drove an entire list of subsequent decisions. Certain applications were simply unavailable on certain platforms. With the rise of UNIX and the independence delivered by open systems, the gulf between hardware platforms narrowed. Hardware is much less of a driving force than it once was.</p> <center> <h5>Figure 7.1. Commoditization of applications.</h5> <p><img border="0" width="321" height="226" src="/books/2/551/1/html/2/images/0130409510/graphics/07fig01.gif" alt="graphics/07fig01.gif"></p> </center> <p>The same is happening with all the layers of this stack. Applications deliver the solution, not the hardware or the operating system. The applications abstract the layers beneath. Few people outside of the IT department are concerned with what lies beneath the applications. Soon, these layers will become as much of a commodity as the hardware. As a particular layer develops into a commodity, it becomes less and less restrictive of the ultimate solution.</p> <p>We discussed the IEBI solution as a solution in which there are three main ingredients : the organization, the Internet, and Business Intelligence (BI). By bringing them all together in one solution, each is transformed by the others. We also discussed the hardware of the Internet. We reviewed the hardware systems and the connections between these systems that create the Internet. Upon which server the application is executing as well as where the server is physically located is irrelevant to the system user . Quite often, a single application will execute across multiple servers.</p> <p>There is a very practical implication to the abstraction of these lower layers. When the hardware, operating system, and even the database become abstracted, we move closer to the goal of establishing a truly integrated information infrastructure. The abstraction of these lower layers removes the barrier that once separated applications running on different platforms. Applications hosted on one server can interact with applications on another. The interfaces to these applications can be tiered together into one common portal. One may argue that this was certainly possible in traditional two- tier client/server technology, where the level of abstraction was not nearly as high up the application technology stack. While this <span>may</span> be true, we must wonder at what cost. It may have been possible, but was it really practical? Was it really seamless? Internet enablement of the application moves the abstraction of the application technology stack to a level at which we can achieve this seamless integration.</p> <p>Let's return to our good friend Billy Boy of Billy Boy Bowling Balls. The good news is that Billy has made some pretty good progress since our last visit. Realizing that his systems were a mess, he hired the world renowned CIO Miles Mody. Figure 7.2 presents the overhaul of Billy Boy Bowling Balls' information infrastructure. As we can see in the figure, all systems were centralized in the IT department. This created an information democracy, enabling individual departments to develop their own customer applications while giving IT the control required to maintain a cohesive enterprise environment.</p> <center> <h5>Figure 7.2. Billy Boy's Internet-enabled information infrastructure.</h5> <p><img border="0" width="500" height="367" src="/books/2/551/1/html/2/images/0130409510/graphics/07fig02.gif" alt="graphics/07fig02.gif"></p> </center> <p>With the exception of the administrative applications running on the mainframe, all of the applications are running on one server in the IT department. It was decided that, at the present time, the administrative mainframe applications were too expensive to replace. They were front-ended with an application server instead. This freed Billy Boy to replace the desktop systems in many departments with lower cost thin clients . This is done only when the older systems have reached the end of their useful life and need replacing. Billy Boy continues to buy laptops when managers or employees need to have a system with them when they leave the office.</p> <p>The same server that supports the Enterprise Resource Planning (ERP) system also supports a class II Operational Data Store (ODS). The ODS provides management with reports that update throughout the day. A separate data warehouse has been implemented that provides long- term analysis and reporting. The data warehouse also supports the organizational data mining and personalization applications.</p> <p>Billy Boy has also tied his numerically controlled devices into the information infrastructure. The devices are able to provide real-time updates on the status of production. This data is included in the daily reports supported by the ODS. It is also incorporated into the data warehouse for the more long-term analyses. Where in the past the production environment was isolated from the rest of the information infrastructure, it has now been fully integrated into the rest of the organization.</p> <p>This is still all hardware. Without the appropriate applications to run in this environment, Billy Boy is not much better off than he was with the old infrastructure. If we succeed in creating a software architecture that maps to this Internet-enabled infrastructure, we will succeed in achieving what we discussed at the beginning of this chapter. The underlying hardware is abstracted. The practical effects of this abstraction means that we have a truly integrated information infrastructure. Systems are able to interact with one another regardless of which server is hosting them.</p> <p>In this section we investigate the software that will provide for this integration. We begin by examining Java. Java is described as "a simple, object-oriented, network savvy, robust, secure, architecture-neutral, portable, high-performance, multithreaded, interpreted, dynamic language."<sup>[2]</sup> It is a language specifically designed for the Internet. Java is more than just a language; it is also a platform for the development of Internet-enabled applications. Developers can now develop true IEBI systems with both the Java language and environment.</p> <blockquote><p><sup>[2]</sup> http://java.sun.com/java2/whatis/</p></blockquote> <p>There are two primary areas where Java's object orientation comes into play. First, the language provides for the abstraction of such BI functions as data mining and On-Line Analytical Processing (OLAP). Through this abstraction, the underlying OLAP, data mining, and database applications become invisible to developers. This enables applications to move more easily from platform to platform. Independent Software Vendors (ISV's) can now develop applications once and have them execute on many different platforms. Second, with Java's object orientation, we can create components that can be later assembled into complete applications. We will examine how these objects, JavaBeans, are used specifically within the area of BI.</p> <p>As we noted, Java is more than just a language. It is a platform for the development of Internet-enabled applications. The Java platform addresses the architectural characteristics of a multitier environment, leveraging the middle tier servers for greater system flexibility and capacity. Another important aspect of this platform is that it is able to address the variety of clients present in an Internet-enabled environment. Using the Java platform, IEBI system developers are able to provide access to wireless devices. Decision makers are now able to access their Decision Support Systems (DSS) virtually anywhere in the world.</p> <p>The Java language and platform provide a means of creating a process, which is only part of the story of Internet-enabling an application. The other part is the data used by the process. The data part of the IEBI solution is the eXtensible Markup Language (XML), which is a mechanism for sharing information between systems across the Internet. In many ways it is just as flexible and architecturally neutral as Java. In Chapter 9 we explore how XML can be used in conjunction with the eXtensible Stylesheet Language (XSL) to distribute strategic information to virtually any client used by a decision maker.</p> <p>In Chapter 8 we discuss the processJavaand in Chapter 9 we discuss dataXML. In Chapter 10 we discuss how the two are brought together through metadata. Metadata flows through the system with the actual data. This mandates that any means of communicating the metadata must be as flexible and platform-independent as the data itself. One solution to the metadata issue is the Common Warehouse Metadata Interchange (CWMI) standard. We will conclude this section with a review of CWMI.</p> </td></tr></table> <hr size="1"> <table width="100%" border="0" cellspacing="0" cellpadding="2"> <tr><td valign="top" height="5"><img src="/books/2/551/1/html/2/images/pixel.gif" width="1" height="5" alt="" border="0"></td></tr> <tr> <td><b><font size="1" color="#0000FF">Team-Fly<img border="0" src="/books/2/551/1/html/2/Fly-Logo.gif" width="81" height="25"></font></b></td> <td valign="top" align="right">     </td> </tr> </table> <table width="100%" border="0" cellspacing="0" cellpadding="0"><tr><td valign="top" align="right">Top</td></tr></table> </td></tr></table></td> <td align="center">  </td> </tr> <tr><td colspan="3" align="center" valign="bottom"> <br><table width="100%"><tr><td height="25" valign="middle" colspan="4" align="center"> </td></tr></table> </td></tr> </table> </blockquote>


Internet-Enabled Business Intelligence
Internet-Enabled Business Intelligence
ISBN: 0130409510
EAN: 2147483647
Year: 2002
Pages: 113

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net