Elasticsearch for Apache Hadoop 1.3.0 M1


See issues on GitHub

Release Notes

New features:

  • write Date/Calendar/Timestamp values as long milliseconds-since-epoch values #79
  • allow option to treat non-existing indices as empty ones #77
  • automatically disable speculative execution #73
  • use bulk API for inserting data #12
  • support for Cascading library #10
  • support shard admin API in ES 0.90.beta2 #2
  • remove jersey dependency #1

Enhancements:

  • Add Hive SerDe support for MAP type #67

Bug fixes:

  • Pig empty Maps #89
  • Hive to Elasticsearch Column names are lost #84
  • Hive Type cast error String to Text #82
  • rename es.column.aliases to something more meaningul #78
  • Unable to run pig job #68
  • number type erasure fix #65
  • Map input is the double of elasticsearch results #63
  • Serialization bug in ESSerDe.hiveToWritable #39
  • ConfigUtils detectHostPortAddress likely broken #5
  • Hive Type cast error String to Text #64
  • Map input is the double of elasticsearch results #62
  • Hive Elasticserach insert a date or timestamp #61
  • Hive writing Error #60
  • Can not create a Path from an empty string (Hive MapRed job with ES) #59
  • ClassCastException VIntWritable and java.lang.Integer #57
  • prefer Map serialization over Iterable #56
  • Duplicate+ Results #55
  • basic update to compile project against CDH4 Hadoop 2/YARN #50
  • `detectHostPortAddress` should return address is not blank #48
  • `detectHostPortAddress` blank check #47
  • Elasticsearch Cluster #46
  • Allow external user mapping to/from Pig #43
  • project cannot be imported into IDEA #42
  • Support for List of Maps (Nested objects) #41
  • change read strategy from pagination to scroll #40
  • Bug in WritableUtils.fromWritable() - incorrect ArrayWritable to List handling #37
  • Import to pig fails when JSON contains an array #36
  • flush index after bulk insert #35
  • Pig storage constructors only take strings #34
  • Invalid test method execution order #32
  • Resolve #30 Embed ES as part of the build and #27 Broken tests #31
  • Embed ES as part of the build #30
  • Publish nightly builds #29
  • Add project to CI server for nightly builds #28
  • Broken tests #27
  • Add assertion that ES is running before tests #25
  • Improve handling of non-existing indecies when loading data #23
  • Expose the ObjectMapper to allow custom JSON serializer/deserializers #22
  • Typo fix in README.md #19
  • consistent configuration load/save across libraries #18
  • fixing code example for writing #17
  • eliminate object saving when performing bulk/buffer updates #16
  • extend Cascading support for Hadoop environment #15
  • Current status of this project? #11
  • incompatible with CDH4 YARN/Hadoop 2 #8
  • Update HTTPClient to HTTPComponents #7
  • Update to Jackson 2.0 #6
  • add jackson serialization for writable objects #3
  • CascadingHadoopTest fails with JsonMappingException: No serializer found for class org.apache.hadoop.io.NullWritable #33
  • Resolve issue37: Bug in WritableUtils.fromWritable() #38
  • Typo fix: temporarely -> temporarily #24

Docs:

  • Fixed the docs to build with docbook #76
  • document existing functionality #71