hive通过外部表读写elasticsearch数据，和读写hbase数据差不多，差别是需要下载elasticsearch-hadoop-hive-6.6.2.jar，然后使用其中的EsStorageHandler；

Connect the massive data storage and deep processing power of Hadoop with the real-time search and analytics of Elasticsearch. The Elasticsearch-Hadoop (ES-Hadoop) connector lets you get quick insight from your big data and makes working in the Hadoop ecosystem even better.

官方：https://www.elastic.co/products/hadoop
下载：https://www.elastic.co/downloads/hadoop

目前最新的版本是6.6.2

# wget https://artifacts.elastic.co/downloads/elasticsearch-hadoop/elasticsearch-hadoop-6.6.2.zip
# unzip elasticsearch-hadoop-6.6.2.zip

使用其中的elasticsearch-hadoop-6.6.2/dist/elasticsearch-hadoop-hive-6.6.2.jar

add jar /path/to/elasticsearch-hadoop-hive-6.6.2.jar;

CREATE EXTERNAL TABLE hive_elasticsearch_table (
id string,
name string,
desc string
)
STORED BY 'org.elasticsearch.hadoop.hive.EsStorageHandler'
TBLPROPERTIES(
'es.nodes' = '$es_server1:9200,$es_server2:9200',
'es.index.auto.create' = 'false',
'es.resource' = 'testdoc/testtype',
'es.read.metadata' = 'true',
'es.mapping.names' = 'id:_metadata._id, name:name, desc:desc');

主要是配置es.nodes、es.resource和es.mapping.names，一个是es服务器地址，一个是index名和type名，一个是hive字段和es字段的一一映射，然后就可以在hive中读写es数据：

select * from hive_elasticsearch_table limit 10;
insert into table hive_elasticsearch_table select '2', 'testname', 'testdesc';

但是这样发现id是被hash过的

+------------------------------+--------------------------------+--------------------------------+--+
| hive_elasticsearch_table.id | hive_elasticsearch_table.name | hive_elasticsearch_table.desc |
+------------------------------+--------------------------------+--------------------------------+--+
| 6mpoc2gBohlnD12tvBoF | testname | testdesc |
+------------------------------+--------------------------------+--------------------------------+--+

还需要再加一个es.mapping.id，定义哪个字段是document的id

CREATE EXTERNAL TABLE hive_elasticsearch_table (
id string,
name string,
desc string
)
STORED BY 'org.elasticsearch.hadoop.hive.EsStorageHandler'
TBLPROPERTIES(
'es.nodes' = '$es_server1:9200,$es_server2:9200',
'es.index.auto.create' = 'false',
'es.resource' = 'testdoc/testtype',
'es.read.metadata' = 'true',
'es.mapping.id' = 'id',
'es.mapping.names' = 'id:_metadata._id, name:name, desc:desc');

这次正常了

+------------------------------+--------------------------------+--------------------------------+--+
| hive_elasticsearch_table.id | hive_elasticsearch_table.name | hive_elasticsearch_table.desc |
+------------------------------+--------------------------------+--------------------------------+--+
| 6mpoc2gBohlnD12tvBoF | testname | testdesc |
| 4 | hello | world |
+------------------------------+--------------------------------+--------------------------------+--+

关于字段类型映射，详见：https://www.elastic.co/guide/en/elasticsearch/hadoop/current/mapping.html