So I tried using multiple keys, but after running make update I believe that beats/libbeat/scripts/generate_template.py did not parse it correctly. I will try again.
So I was able to figure out how to add more sub-fields except that the way that it works is it creates nested fields under the _default mapping.
I usually like to use document _types to pull apart.. well document types. Is there a reason you have designed it this way? I can see that you are going to start using the concept of modules. For example an nginx filebeat module that I assume would create fields like this:
_index: filebeat-*
_type: filebeat
nginx.access.request.method: GET
To me it makes sense to have it be like:
_index: filebeat-*
_type: nginx-access
request.method: GET
I just think more deeply nested field names are harder to query?
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.