To give a follow up, I've setup a docker compose with node-red, timescaledb and grafana on the edge (on a synology NAS) I hope this gives a prove of value but also points out the disadvantages like accessability, scalability and maintenance. We have one machine sending it's data to HiveMQ broker as well right now. Not sure how to I should make use of this best tho. Any way to connect this to a historian/database in a easy way? (take the full UNS tree and directly dump it into a corresponding database setup keeping the topic structure in mind)