Fixing BulkAll indefinitely hanging when failing (#5077) (#5091)

stevejgordon · rmadisonhaynie · web-flow · commit fc84ad2c2f86 · 2020-11-12T11:33:42.000Z
Ensure BulkAll() is aborted properly if the whole bulk request keeps returning a bad statuscode and ContinueOnDroppedDocuments is true. This was introduced in #4014 when trying to address a similar issue with backOffTries being ignored. Co-authored-by: R. Madison Haynie <rmadisonhaynie@yahoo.com>
diff --git a/src/Nest/Document/Multiple/BulkAll/BulkAllObservable.cs b/src/Nest/Document/Multiple/BulkAll/BulkAllObservable.cs
@@ -202,7 +202,7 @@ private async Task<BulkAllResponse> HandleBulkRequest(IList<T> buffer, long page
 
 			void ThrowOnExhaustedRetries()
 			{
-				if (_partitionedBulkRequest.ContinueAfterDroppedDocuments || backOffRetries < _backOffRetries) return;
+				if (backOffRetries < _backOffRetries) return;
 
 				throw ThrowOnBadBulk(response,
 					$"{nameof(BulkAll)} halted after {nameof(PipelineFailure)}.{reason} from _bulk and exhausting retries ({backOffRetries})"

Original file line number	Diff line number	Diff line change
`@@ -202,7 +202,7 @@ private async Task<BulkAllResponse> HandleBulkRequest(IList<T> buffer, long page`
`202`	`202`
`203`	`203`	`void ThrowOnExhaustedRetries()`
`204`	`204`	`{`
`205`		`- if (_partitionedBulkRequest.ContinueAfterDroppedDocuments \|\| backOffRetries < _backOffRetries) return;`
	`205`	`+ if (backOffRetries < _backOffRetries) return;`
`206`	`206`
`207`	`207`	`throw ThrowOnBadBulk(response,`
`208`	`208`	`$"{nameof(BulkAll)} halted after {nameof(PipelineFailure)}.{reason} from _bulk and exhausting retries ({backOffRetries})"`