Elasticsearch for Apache Hadoop version 6.3.0edit

June 13, 2018

Enhancementsedit

Rest
  • Add additional error message if all shards are unavailable #1129
  • Add switchable findPartitions behavior. #1108
Build
  • Overhaul the build system to be modular. #1120

Bug Fixesedit

Rest
  • HttpStatus.GATEWAY_TIMEOUT should be 504, is 50 #1147
  • Scores are null when reading metadata from scroll data #1137
  • can’t set "http.connection.timeout" in HttpClientParams. #1124
Serialization
  • Setting es.read.field.include breaks mapping discovery for fields nested deeper than two levels #1139
  • Array include depths not being respected on singleton value fields. #1134
  • NullPointerException when setting “es.read.field.as.array.include” option #928
  • SparkSql scala.MatchError when inserting single-value object in nested type #808
Hive
  • getting a wrong result when run a hive-sql that contains union-all opeartor. #1084
  • Fixed resourceRead field from settings to read dirty data when contains the union-all opeartor in hive-sql. #1085
Spark
  • DefaultSource wrongly parses es.resource (spark2.2) #1082
  • Unable to overwrite dataframe when documents have parent #1030

Documentationedit

  • Update arch.adoc: correct backtick quotes #1136
  • Update configuration.adoc #1126
  • Update spark.adoc for Windows #1121