Enterprise Search is one of the topics that required in every organization but sometimes it doesn’t get enough resources. I don’t mean CPU, memory, etc. but human resources during the plan, deploy and configure the search infrastructure and architecture.
Let me summarize some key components that you need to plan if you’d like to deploy an Enterprise Search for your SharePoint.
As Seach plays different roles in every different organization, there is necessary to plan deeply and deploy carefully for an effective implementation. The following figure shows the key components of the Enterpise Search:
Enterprise Search key components:
  • Physical architecture:
    • Index server: Index server is responsible for crawling and indexing the various contents. Its general tasks are:
      • Connects to a content source;
      • Readsg the content and metadata;
      • Places the metadata in the SQL property store;
      • Places the content in the full text index.
    • Query server: Query server accepts search queries from the users, builds return set and return the results. This is the typical interface where users meet the SharePoint’s query engine.
    • Shared Services Providers (SSPs): The SSPs are responsible, among others, for deploying and configuring the SharePoint Enterprise Search and making its functionalities available for the various applications and sites.
  • Logical architecture:
    • Content sources: SharePoint Server 2007 can index various content sources:
      • SharePoint sites
      • File shares: documents (Office formats, PDF, TXT, etc.) stored in shared folders of the file system;
      • Business Data: data stored by external Line-of-Business systems in various databases (eg. MS SQL, Oracle, etc.) or accessed via Web Services;
      • Websites: any type of web content inside or outside of the organization;
      • Exchange Public Folders.
    • Search scopes: Search scopes refine the queries. Scope rules can be built by web addresses, property queries and content sources, or can be referred to the all indexed contents.
    • Federation: With federation, SharePoint has the capabilities of querying content sources managed by some other crawling engine. It can conserve the resources of crawling and indexing, moreover can include content that cannot be crawled by SharePoint.
    • Keywords and Best Bets: Keywords mark specific items as more relevant, therefore they show up more prominently in the search results. By best bets, relevant items and keywords can be chosen and highlighted for a subject.
  • User experience:
    • Search pages
    • Navigation
    • Tag cloud
    • etc.