I am using the NEST API 6.5.1.
When indexing data I need to separate the errors for operations that should be automatically retried and the errors for operations that should not be retried. I am using the BulkDescriptor.IndexMany method and the ElasticClient.BulkAsync method. For example:
If an "Invalid index name" error is returned then the error should be logged but the operation not automatically retried.
If a network error is return eg a sniffing error then index operation should be retried later when network connection is restored.
Currently I am examining the IBulkResponse.Items collection to determine the error.
I cannot determine a clear way separating the two types of error. Can anyone provide assistance?