I'm evaluating the elastic stack as the successor tool for the analysis of production data.
We are getting documents from our production systems that are annotated with some kind of ID in string format, lets say "Q39100232". In our current system we are maintaining a table containing this IDs with a multi language description field.
So depending on language config Q39100232 maps to the english string "RPM xxx" or german string "Drehzahl xxx" and so on. (10 Languages)
Currently we have the ability to filter the our events based on the ID. The input box you enter the filter ID into supports autocomplete with the description texts as well. So all in all your able use the input box to search and filter at the same time.
I've indexed the documents with their corresponding IDs as an keyword in elasticsearch and I'm able to filter the documents in Kibana based on the ID string Q39100232 like a charm.
My solution would be:
At index time I can make a request to our current system an request all translations for Q39100232 and add them as PID.EN, PID.DE, PID.CZ etc to the document. I would be able to search all languages like a charm. But someday the description texts change and I would need to update all documents containing the changed text.
Has anyone an idea solve this in a more elegant way?