Uncertain about denormalisation of the Elasticsearch indices with regards to best practices
Referring to: Search Behaviour Wiki Page
I am unsure how the indices were denormalized in R1, however it is best practice to denormalise the data when dealing with Elasticsearch. Normalised data typically requires joins which either need to be conducted on the database side or client side.
Where could I find the old and new Elasticsearch index mappings so I can understand the difference between them and provide feedback on possible denormalized/normalised capabilities and limitations?
Furthermore, I would be also interested in seeing if there are any documented "search use cases" to understand how they are currently being achieved with denormalized and understand how that changes with regards to normalised data. Where can I find such documentation?