search

jprante / elasticsearch-jdbc

JDBC importer for Elasticsearch
Apache License 2.0
2.84k stars 710 forks source link

Help!does jdbc-importer support ES template? #861

Open Judyccb opened 8 years ago

Judyccb commented 8 years ago

My ES version is 1.4.4,jdbc-importer version is 1.7.3 My problem is: I do not want to define "index_settings" or "type_mapping" in JDBC importer definition file,I want to define them in ES template.does jdbc-importer 1.7.3 support it?

jprante commented 8 years ago

This feature is outside of JDBC importer. Create index template in ES, and use it from JDBC importer by

select myindexname as _index, ... from table ...

where myindexname is the index name that matches the pattern configured in the index template.

Judyccb commented 8 years ago

Thanks for your help. 1、where should I define the command "select myindexname as _index, ... from table ..."?in the JDBC importer definition file? What I want to do is import data of table "aaa" from Oracle to ES, Do you mean I should change "select * from aaa " in JDBC importer definition file into " select abc-index as _index from （select * from aaa)"?. I'm confused. 2、Could you give me an example? Thanks for your patience again!

jprante commented 8 years ago

Of course you must define the SQL statement of the JDBC importer.

select * from aaa is a naive example SQL which works, but it does not address Elasticsearch index, type, or id. It is just a simple test. The asterisk * is not suitable for Elasticsearch field names either.

To create Elasticsearch documents, you need something like select myindexname as _index, mytypename as _type, mydocumentid as _id, myvalue1 as fieldname, myvalue2 ... from aaa where ...

The pseudo column names beginning with _ are documented at https://github.com/jprante/elasticsearch-jdbc

You can freely choose myindexname, mytypename, etc. as you like, as literals or values from DB.