The search index is a set of files that are stored in separate folders on a server. The Content Processing Component processes items provisioned by the Crawl Components, maps crawled properties to managed properties, and formats these as artifacts that can be stored on the search index. The indexes can include:
- Full-text indexes.
- Indexes of the managed properties (marked as retrievable or queryable).
- An index for attribute vectors
- Numeric indexes.
Because the Search Index can be a very large file, you can divide it into index partitions to improve performance and management. An index partition is a logical portion of the entire search index. The index comprises the Index Component, index partitions, and index replicas, which are described as follows:
- Index Component. This is used both in the content ingestion process and in the content querying process. For the former, the Indexing Component receives items from the Content Processing Component and writes them to the index file. For the latter, it receives queries from the Query Processing Component and returns the result sets that are applicable to the query.The Indexing Component is also responsible for much of the index content management. For example, it will physically reorder index content if changes to the index architecture are triggered by the Search Administration Component.
- Index Partition. This is a logical portion of the entire search index; the index is the aggregated result of all of the index partitions.
- Index Replica. This is a physical copy of an index partition. Replicas can be either a Primary Replica or a Secondary Replica. The Primary Replica is contacted by the Content Processing Component to write new items to an Index Partition. Secondary Replicas are read-only copies of the same data, which are used to provision results. You can scale you search index in two ways:
- You can add Index partitions to manage increasing search content volume. For example, in a farm with three index partitions, each index partition contains one-third of the entire search index.
- You can add Index replicas in an index partitions to manage high query loads or to provide increased fault tolerance. Each index partition has one or more index replicas. For example, in a farm with one index partition that contains three index replicas, each index replica serves one-third of the total queries