Team at a place I work at picked OpenSearch as a primary (and only) database for the product (collaborative CRUD). It feels wrong to me, as I’ve seen issues around data loss, consistency problems, performance issues in the articles/tests in the past [0]. But there’s not much info about the recent versions of ElasticSearch/OpenSearch as a main DB, so I decided to ask here - does anyone have experience using ES/OS as a primary datastore (at scale) or has some useful references or suggestions for evaluating such decision? Thanks!

[0] https://aphyr.com/posts/323-jepsen-elasticsearch-1-5-0