Pending Elasticsearch streamed works may fail when stopping Hibernate Search and using the drop-and-create-and-drop strategy

Description

Currently, we only wait for those works to finish when closing the ElasticsearchService, which I think happens after the index managers have been closed.
Since the index managers are responsible for dropping the indexes if necessary, this can result in very strange behavior: async works failing because the index doesn't exist anymore.

Setting the priority to "minor" since the "drop-and-create-and-drop" strategy is not likely to be used for anything else than tests.

Environment

None

Status

Assignee

Yoann Rodière

Reporter

Yoann Rodière

Labels

None

Suitable for new contributors

None

Feedback Requested

None

Components

Fix versions

Affects versions

5.8.2.Final

Priority

Minor
Configure