[ML] Make use of allow_partial_search_results
on search requests of data extractors
#40793
Labels
allow_partial_search_results
on search requests of data extractors
#40793
When datafeed data extractors run searches they expect no errors on the search responses. Currently, we check for these errors manually. However, there is a feature that allows searches with partial results to fail with an exception instead of responding. We should revisit and simplify data extractors by setting
allow_partial_search_results
tofalse
in the relevant search requests. Note, an additional benefit doing this is that when the search fails any scroll context gets cleared automatically, reducing the chance of data extractors leaking scroll contexts (see #40772).The text was updated successfully, but these errors were encountered: