Elasticsearch for Apache Hadoop 1.3.0 M2


See issues on GitHub

Release Notes

New features:

  • Write RawJsonWritable #126
  • Fix for issue #121. #123
  • Integer data mapping issues #121
  • retry functionality #114
  • add security support #111
  • add parent/child support #107
  • dynamic concurrent writes #106
  • allow ES mapping/directives to be embedded before a job execution #103
  • upgrade to Apache Hive 0.12 #101
  • upgrade to Hadoop 2.2.0 #100
  • add support for Pig 0.12 #98
  • Error reading nested documents from ES #97
  • add ordered MapWritable compatible implementation #93
  • support for routing keys #85
  • Nested documents #81
  • Indexing JSON #75
  • Index aware document writes #69
  • Add support for query fields= #54
  • allow raw json to be used without parsing #9
  • HDFS support for ES 1.0 snapshot/restore feature #72
  • add parent/child index feature #66

Bug fixes:

  • Documents containing subdocuments are not handled correctly by pig interface - bug in projection logic in PigUtils.java? [OPEN] #117
  • Multifield Mapping wihout default field causes a Nullpointerexception #132
  • User _source instead of fields parameter to support nested objects. #130
  • Fields in tuples are generated out of order in pig interface #119
  • wrong query generated by hive #118
  • If the date data in ES format is 2013-12-01 12:12:12.876 will report an error #116
  • HiveFieldExtractor null pointer exception #115
  • CascadingValueWriter should also handle hadoop value types and not just jdk value types #110
  • Generated query URLs are malformed if the source resource URL does not have a query portion #104
  • No Class def error when running CDH3 #95
  • Pig Map support in ESStorage #92
  • Pig Storage index name with "_" #91
  • RestClient might send an invalid/empty bulk request resulting in errors #90
  • org.elasticsearch.hadoop.rest.Resource parses some valid index-names wrongly #80
  • Case sensitive mapping Hive #58
  • org.elasticsearch.hadoop.rest.Resource: parsing of resource fails when the search string contains '_' #102
  • Providing the document ID #112
  • Cascading tap should use the field declaration as a selector #108
  • document fix #105
  • A null pointer exception is thrown when a node in a cluster does not have a defined "http_address" #99
  • set execute permission to the gradlew script #94
  • no support for multi_field type #88
  • allow multiple hosts (for resilience/load balancing) in es.resource #74
  • Add support for explicit document _ids when Hadoop key is Text #53
  • some exceptions easy to fix #52