Hello all, hope you can help me.
Imagine I have an Elasticsearch index "sale-merged" and 2 different types (_traffic, _bi-merged) within one index. Types store simillar data and the mappings are amost the same, just _bi-merged has additional fields.
In ES I can run a query to specify from which _type to retrieve documents: { "query": { "match": { "_type": "traffic" } } }
But at the stage when I configure the index to be used by default in Kibana, I see only documents that are stored under _bi-merged type, but no documents in _traffic type - http://prntscr.com/btaahe
The thing I want to achieve is to apply relation-database approach - where index is a database and _type is a table.
In the end, I need to be able to work with documents from both types in Kibana and Timelion
Well, we proceeded with perhaps not the best decision to build BI system on top of Elastic, and now this is more of a patching process.
Actually, the issue can be resolved if when configuring the index in Kibana, I set the time field that exists in the mapping of both types - http://prntscr.com/bu9kwq
Thus, I can use the filter '-type "traffic"' to exlude documents of an another type, but does not seem very consistent - http://prntscr.com/bu9lax.
Thanks for the article @warkolm, I guess in our use case, type use will be convinient.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.