Search Service Application: Architecture in one page

Search service application consists of the following pieces:

1. Administration interface (_admin/ServiceApplications.aspx)

2. Crawl component

3. Databases (Administration, crawl, property)

4. Index partition (Query component) 

Crawler - a component that browses the contents automatically on a regular basis. The goal of the crawling process is to provide up-to-date data from the data store to the indexer(s).

Indexer -a component that collects and stores crawled data. The goal is to make data available for fast and precise information retrieval during the queries.

Query -  a component that provides UI for entering the user queries, presents the result set to the end users. It communicates with the Indexer component directly to get the result.

(Real World SharePoint 2010: Indispensable Experiences from 22 MVPs)

SharePoint 2007 defined 2 search roles: Query and Index.

With modularity of SharePoint 2010, these functions are defined as components.

The Index role(2007) has been renamed as the crawl component (2010).

In 2010 you have 2 components:

 - Crawl component;

 - Query component.

In 2010 you can make Search fault tolerance by having several crawl and query components (with option "Failover-only Query Component").

To run a Query component on a server, the service "Search Query and Site Settings Service" should be started. (refer to Service Application : Architecture in one picture)