BulkAll doesn't retry 429s when ThrowExceptions is enabled

[nobowned]

NEST/Elasticsearch.Net version: 7.10.0

Elasticsearch version: 7.10.0

.NET runtime version: .net 6.0

Operating system version: Windows 10.0.19044 Build 19044

Description of the problem including expected versus actual behavior: BulkAll won't retry 429s if ThrowExceptions is enabled.

Steps to reproduce:

1. Turn on ThrowExceptions.
2. Call BulkAll with BackOffRetries set to something greater than 0.
3. Call Wait on the BulkAllObservable returned by BulkAll.
4. Ensure a 429 will occur. This can be done by bulk indexing more documents than your cluster can handle.

Expected behavior 429s are retried with ThrowExceptions enabled.

[nobowned]

A quick fix could be to disable ThrowExceptions on the BulkAsync call performed by BulkAllObservable:

https://github.com/nobowned/elasticsearch-net/blob/f6bbf21e81c22d9fe160725f87e16ab2ed0cdbdb/src/Elastic.Clients.Elasticsearch/Helpers/BulkAllObservable.cs#L120-L124

I think this should be fine since BulkAllObservable will throw it's own exceptions if something goes wrong anyways.. regardless of the "ThrowExceptions" setting.

[stevejgordon]

Thanks for raising this @nobowned. I'll need to spend some time on this to reproduce and determine the best option for this.