:Summary index

Introduction to Summary Index

A summary index stores all the information related to a single document. It uses a document ID (docID) to locate this information and provides a service for storing and retrieving document summaries. The structure of a summary index is similar to that of a forward index, but they serve different purposes.

Sample code for configuring a summary index

"summarys":
{
	"summary_fields":["id", "company_id", "subject", "cat_id"],
	"compress":false,
  "parameter" : {
      "compress_type" : "uniq|equal",    
      "file_compress" : "simple_compress1"
  } 
}

• The summary_fields parameter specifies the fields that you want to store in your summary index. You can specify fields of all data types for this parameter.

• The compress parameter specifies whether to compress your summary index using zlib. The valid values are true and false. The value true indicates to compress your summary index, and the value false indicates to not compress your summary index. The default value is false.

• When you specify a field of the TIMESTAMP data type as a summary index, the field is stored as an attribute. This way, the time format is processed by the system in a centralized manner if the default time zone is used.

• You can use the parameter field to configure specific parameters. OpenSearch Vector Search Edition V3.9.1 and later versions support the parameter field. The following parameters are supported:

• compress_type: The codec. For an explanation of uniq and equal, see the Fields section.

• file_compress: indicates to compress files. You must specify the alias of a compressed file for the file_compress parameter after you specify the alias in the schema.json file. The method specified by the file_compress parameter provides better performance for compression than the method specified by the compress parameter. We recommend that you set the value of the compress parameter to false and configure the file_compress parameter to provide better performance for compression.