Using the Taxonomy
Now that we've seen the categories and their relationships with one another, we are in a better position to understand vendor technology and to position it in terms of what it offers to an integration project. Most products or product families span the categories. However, individual components often fit within one category, therefore making comparison relatively straightforward.
As an example of how the taxonomy can be used, we'll look at the MQSeries product family from IBM and position its components. MQSeries itself is a data-exchange product. It provides traditional message-queuing and publish/subscribe capabilities. It also provides a robust message-queuing function with transactional capabilities that are used in conjunction with many other products. The MQSeries Integrator product provides full message-brokering capabilities, including transformation and routing. When used together, these two products provide a very complete set of data exchange capabilities.
However, MQSeries extends beyond the basic data exchange category in two ways. First, in its latest versions it provides application-level transactions, although the facilities are less comprehensive than those of dedicated products, such as Tuxedo. Secondly, some basic data-integration capabilities have been added to MQSeries Integrator Version 2. The third member of the family is MQSeries Workflow, which provides traditional work-flow management capabilities and uses the other MQSeries products for integration.
Now that we know where the members of the MQSeries family fit in the taxonomy, we can address questions about specific application-integration projects. For example, suppose we need to implement a data-integration scheme that transfers very large files among several remote sites. In addition to MQSeries family products, what else would we need in order to implement this solution? Although MQSeries provides just about everything we need in the data exchange category (aside from security), it most obviously does not provide the batch data-integration facilities we're going to need. We'll need an additional product to perform the file transfers using MQSeries. In addition, if our data is going to flow over external networks, we may decide that we need encryption or digital signature as well as some form of authentication control. These might be included in the file transfer product that we purchase, or we may need additional products.
Hopefully you can see from this simple example how the taxonomy can help you to organize your thoughts about the set of products that are needed to implement the capabilities required to support an eAI project. Where multiple product choices are possible, you can perform an analysis for each and compare the results. For example, a number of robust file-transfer products do not use MQSeries. You might choose among the possible alternative solutions by analyzing these products and comparing the results. Knowing where each product fits within the taxonomy makes it much easier to compare potential solutions to a specific problem. When you've categorized a vendor's offerings, you can compare the features of specific products or product combinations within the categories to help you decide which will be more appropriate