Is there a way to get the original Elasticsearch response in Kiabana instead of the tabified response? I found a thread where it was stated you could set your response handler to 'none' and that would return the original Elasticsearch response, but this was for version 6.1 of Kibana and I'm using version 7.5. I've tried setting my response handler to responseHandler: 'none', but I'm still getting back a tabify response. Is it possible to get an non-tabify response with version 7.5?
in what context are you trying to achieve this ? i am guessing you are trying to add a new visualization type ? no, that is not possible, we changed the architecture of visualizations trying to decouple them from data fetching. For this reasons visualizations are not given any raw es responses but rather a datatable which we use as a common data model. There is also no information about the aggregations you ran and so on, as the idea is the visualization should not care where the data came from, which allows us to reuse same visualizations for different data fetching methods.
what you most likely want to do is implement your own embeddable. Where visualization is responsible only for rendering and should not care where the data comes from, embeddable does the data fetching and gives the data to visualization (or you don't need visualization at all, embeddable does rendering as well. However that will not allow you to reuse the default editor, you will most likely end up writing your own kibana app which will serve as the editor for your embeddable.
Thank you for the detailed response. You are correct in that I'm writing my own visualization. I use bucket aggregations and I'm trying to find ways around the issue of the aggregations not being accurate at times. What I found was the count that is returned for the __other__ bucket with the tabify response can be a lot different than the count for the sum_other_doc_count parameter that gets returned with the raw ES response. I've found the sum_other_doc_count to be more accurate and I wanted to use that count instead of the __other__ bucket count for calculating the percentages. I'll look into the embeddable. Thanks.
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.