Ive got two tables of URLs, one of them counts the number of visits for the root path of a URL and the other is a list of all URLs visited. The field, url.path, is indexed as both a text and keyword field. I would like to be able to filter by a value on the root path table and the all urls table show everything under the root path.
As an example, below is my root paths table.
I would like to be able to filter by /healthandwellness, which in turn would display the following in my all URLs table:
Right now, it doesn't work that way, because they're both looking at the keyword field. I guess the better question is, can I create a table that aggregates on a non-keyword? I realize I can use the search bar and enter url: \/healthandwellness, but I'm trying to get this to work for non-technical users.
What I would do, either at ingest time or with a scripted field would be to create a field that only contains the root path, and then use that to display in the first table and filtering with + and - signs will work properly.
As shown in the images of the original post, I already have a field that contains only the root paths. The field is stored as both text and keyword data types, but Kibana will only let me create a table based on the keyword data type. Meaning that filtering on that table's value will only return exact matches.
You can only visualize keyword fields, but you can add a filter using the full query DSL- so if you have a text mapping and a keyword mapping for these paths, you can filter by the text and visualize the keywords.
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.
