Architecture of Indexing and Search

Because the Search component is actually the only element of the MOSS infrastructure that tends to span just about any physical piece of the environment, it is also the most challenging to set up and to configure. A simple introduction to its architecture and to the inner workings of the system will potentially avoid future setup nightmares and configuration pitfalls.

To start off, see Figure 27-9 for a graphical representation of communication paths between some of the core features and elements of Search. In the middle, we have the core elements, the Index engine, and the query engine behind Microsoft Search. The index engine retrieves search configuration data from the database and is responsible for crawling the data sources and compiling the index. Query Engine, on the other hand, only works with the built index (as well as some additional configuration data) to serve the Search results.

Warning

Contrary to popular opinion, MOSS Search does not perform a full-text search directly against the data in the database. The gatherer part of the Index engine actually walks through every known piece of content against a WebFront End (WFE) server using standard HTTP protocol, discovering the content before deciding whether it fits the criteria of being indexed.

Figure 27-9. Core elements of Search technology