After loading lots of data and lots of fields on the Elastic Stack it starts becoming very difficult for users (especially non-IT) to keep track of what each field actually represents. Essentially what's needed is a Data Dictionary.
This could be totally separate from the Elastic Stack but it would be awesome if the stack provided it.
Maybe there's already some way to "hack" something like that in? If not, what solutions do people use to solve this problem?
I'd be interested in how you think this might be possible for the stack to provide it? It's not a bad idea, but ultimately ES relies on the you, the user, to define things based on your use case and domain knowledge.
I feel the most natural place for this would be mappings. One could provide a description field along with the type or whatever else they want to specify. This could be then retrieved by Kibana and displayed in a popup over a question mark or something next to the field.
This could probably be implemented as a plugin too by using an ES index for storing the description data and hacking access to them on various parts of Kibana.
I found a few related issues that ask for the type's
_meta field to be added on the field level but they all seem to have been rejected one way or another:
Ahh ok that makes sense.
I guess in that case it's really just another field:value, and it'd be up to the client (ie KB) to correctly render it.
Uhmm which one makes sense? The mappings support, the plugin or _meta for fields?
I started looking around for info regarding Kibana plugins and it seems that messing around with the Discover app's UI will have some caveats as mentioned in this thread. I'm still not sure though, so I'll probably have to look more into how they work.
Your general requirement.
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.