This chapter is from the book
This chapter is from the book
This chapter is from the book
3.6 Conclusion
Today's enterprise application landscapes are characterized by a variety of different technologies and concepts for distribution. On one hand, this variety arises within the enterprise organization itself for historical reasons, personal preferences of different people, and the dynamics of acquisitions and mergers. As a matter of fact, many redundant concepts exist within the same organizational unit. On the other hand, complementary concepts and technologies also exist. Due to the requirements of different types of distribution problems that coexist in one corporation, different solutions arise as well.
A modern architecture must be able to embrace all these technologies and concepts. Heterogeneity—including heterogeneity of middleware—must be understood as a fundamental fact that cannot be fought but instead must be managed. Furthermore, an architecture must accommodate frequent changes of the underlying distribution infrastructure. As a matter of fact, the lifecycles of today's infrastructure products are largely incompatible with the lifecycles of enterprise applications. Thus, you must protect the assets of an existing application landscape and simultaneously take advantage of the latest infrastructure products.
In this chapter, we have discussed the necessity of carefully choosing the right approach to integrating two distributed software components. Among other issues, you must decide on the appropriate communication infrastructure, synchrony, call semantics, usage of an intermediary, and object-oriented versus data-centric interfaces. All these decisions impact the coupling of the two systems.
References
Braswell, Byron, George Forshay, and Juan Manuel Martinez. IBM Web-to-Host Integration Solutions, 4th ed. IBM Redbook SG24-5237-03, 2002. [Bras2002]
Long, Rick, Jouko J_tti, Robert Hain, Niel Kenyon, Martin Owens, and Andr Schoeman. IMS e-business Connect Using the IMS Connectors. IBM Redbook SG24-5427-00, 1999. [Lon1999]
Tanenbaum, Andrew S. Computer Networks, 4th ed. Prentice-Hall, 2003. [Tan2003]
Tanenbaum, Andrew S. and Maarten van Steen. Distributed Systems: Principles and Paradigms. Prentice-Hall, 2002. [Tan2002]
Coulouris, George, J. Dollimore, and T. Kindberg. Distributed Systems Concepts and Design, 3rd ed. Addison-Wesley, 2001. [Cou2001]
Reisig, Wolfgang. A Primer in Petri Net Design. New York: Springer Compass International, 1992. [Rei1992]
URLs
Footnotes
[1] Often referred to as TP monitors, TPMs, or TX monitors.