Skip to content

Fill bulk errors with nested error causes #1115

@ebuildy

Description

@ebuildy

What kind an issue is this?

  • Feature Request

Issue description

Add elasticsearch error causes in error reporting in order to help issue resolution. See the todo at https://github.com/elastic/elasticsearch-hadoop/blob/master/mr/src/main/java/org/elasticsearch/hadoop/rest/bulk/BulkProcessor.java#L259.

Or maybe, append all causes with the ErrorExtractor class.

Steps to reproduce

Insert documents from Spark in elasticsearch index that will not respect the mapping, this will give bulk error:

"index": {
        "_index": "consolidation-test",
        "_type": "all",
        "_id": "Qagq-2EBTt5y_FPQwFd7",
        "status": 400,
        "error": {
          "type": "mapper_parsing_exception",
          "reason": "failed to parse",
          "caused_by": {
            "type": "illegal_argument_exception",
            "reason": "Could not convert [version.index] to boolean",
            "caused_by": {
              "type": "illegal_argument_exception",
              "reason": "Failed to parse value [not_analyzed] as only [true] or [false] are allowed."
            }
          }
        }
      }

But elasticsearch4hadoop will only report "failed to parse", that dont help develop to understand whats really happened under the hood.

Version Info

OS: CentOS 7
JVM : 1.8
Hadoop/Spark: Spark 1.6.3
ES-Hadoop : 6.2.2
ES : 6.2.2

Feature description

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions