In the previous chapter we looked (albeit briefly) at the issues involved in authenticating users for our applications, through Windows NT and Internet Information Server (IIS). Now we can move on to see how this affects the design and construction of a DNA-based application. DNA applications follow a structured three-tier approach to design and implementation. But although this is considerably simpler than the multi-layer designs of some traditional applications, it still requires some careful planning as to how each tier will interface with those above and below it.
In this chapter, we'll finish off the 'teaching' section of this book with a look at a whole range of issues that affect how successfully and reliably your applications work when you move them from that development server out into the 'real world'.
The issues fall into three main areas, mirroring the three parts of a DNA application that we basically ignored in the previous chapter. The areas are Microsoft Transaction Server (MTS), Message Queue Server (MSMQ), and your chosen Data Source. Before we go into these in detail, however, we'll spend a little time looking at the way that communication between them and other components works—with the emphasis on the way that each one authenticates the 'user' that invokes them.
So, this chapter covers:
We'll begin with an overview of the authentication process, to see where we need to configure each layer and how we should approach it.