When you run the Initially Index Smart Search Data batch job on a base object, the Solr server indexes the data in the searchable fields. The job then adds the indexed data to all the collections that represent the business entities to which the searchable fields belong. If a collection is too large, you can split the collection into one or more shards. Shards are the logical pieces of a collection split over multiple nodes. When you perform a search, the Solr server reads the collections and returns the matching fields.

The Initially Index Smart Search Data batch job indexes the records asynchronously and reports successful completion after the job queues the indexing request for all the records. A search request can show the indexed records only after the successful completion of the index request, which might take a few minutes.