Currently, there is an API call that forces Bitbucket to attempt a re-index of all repositories that is available by making a POST request to "/rest/indexing/latest/sync".

However, this reindex is performed using the indexes that already exist, meaning that the only way to clear out corrupted/bad index entries is to fully remove all indexes from the filesystem while Bitbucket is offline (as per Resolution #2 in the following knowledge article: Rebuild the Elasticsearch Indexes for Bitbucket Data Center - Atlassian Support).

This request is to implement a REST API call that, when invoked, will initiate a full background reindex during runtime that will clear out any existing index entries and then begin the index sequence from scratch for each repository.