Honestly folks choose elasticsearch only in part for Trie over B+tree. It provides a fair amount of magic wrt to appropriate defaults for dealing with large datasets in a distributed store.
I think a lot of folks would be better off with RMDB, but if you barely know SQL and spend most of your time making UIs, you’re lucky to have the breadth of know-how to configure Postgres the right way (no offense intended to frontend developers).
Of course, Elasticsearch’s magic defaults expectations may come back to bite you later on when you’re using it OOTB this way, but it’s hard to argue with throwing a ton of data in it and then -POOF- you have highly performant replicated queries, with views that are written in your “home” programming language, without even really necessarily understanding what your schema was to start with (yikes, but also, I get it).
I think a lot of folks would be better off with RMDB, but if you barely know SQL and spend most of your time making UIs, you’re lucky to have the breadth of know-how to configure Postgres the right way (no offense intended to frontend developers).
Of course, Elasticsearch’s magic defaults expectations may come back to bite you later on when you’re using it OOTB this way, but it’s hard to argue with throwing a ton of data in it and then -POOF- you have highly performant replicated queries, with views that are written in your “home” programming language, without even really necessarily understanding what your schema was to start with (yikes, but also, I get it).