I use es_hadoop_hive to save data into elasticsearch. But I have a field which I formatted as string in hive but the mapping is geo_point in ES. How can save the data such that the mapping auto generated by hive does not override the geo_point mapping already created?
Or is there a way to convert a field type from String to Geo_point when visualizing in Kibana?
It won't over write what is in ES. Is that what you are seeing?
No, you need to map it correctly in ES.
I did the mapping correctly in ES but anytime I insert into the Hive table that connects to ES, another type is created with the field name as the same but fieldtype as string.
Maybe if you post your mappings we can comment further.
I have the same problem
I need to convert a string coordinates to geo_point type to visualize in kibana correctly.
I've tried to convert this field in the hive script, it's possible?
CREATE EXTERNAL TABLE hive_to_es_table
(
id string,
geo_coordinates string
)
STORED BY 'org.elasticsearch.hadoop.hive.EsStorageHandler'
TBLPROPERTIES('es.resource' = 'my_index/my_type',
'es.mapping.id' = 'id',
'es.mapping.names' = 'geo_coordinates:@geo_point',
'es.nodes' = 'minode.corp');
I have not gotten it.
Can anybody help me?
Thanks
You should start your own thread over in the Hadoop category 