If you modify any configuration using the Provisioning tool, the Elasticsearch indexes are deleted.

When you run the Initially Index Smart Search Data job, the Elasticsearch indexes are deleted.

After applying a filter, the filter count for referenced columns is not updated.

When a parent business entity is also a one-to-one child of another business entity and both have some searchable fields in common, the Initially Index Smart Search Data job fails.

When you use accented characters in the search string, records are not found.

When a base object has a large number of root records, indexing lookup tables might generate an out of memory error.

After you change the base object class code, search results display duplicate record entries.

When more than 50 searchable fields are configured for a business entity, indexing fails.

When you perform a search, the MDM Hub ignores common words such as "and", "an", and "is". These are stop words, which are not filtered out by the suggester.

If you make changes to a lookup for a record, reindexing for the record fails.

When records contain dependent lookups, the load update process takes longer than expected to index the records.

Duplicate records appear in the search results even after the duplicate records are merged.

When you have two process servers in a node with Smart search enabled, search fails with errors.

A smart search query that includes AND or OR in the search string fails.

When search is enabled, if you search for a record after merging some records, duplicate records might appear in the search results page.