I am wondering why there are no solutions to data warehousing that work in a p2p manner rather than relying on centralised servers. The obvious upside to have a cloud-hosted/on-premise DW is that you have a single source of truth from which all further analysis is derived. On top, there is also a dedicated team that does query optimization, data integration and so on.
But especially for heavily regulated or understaffed companies, where development cycles are long, this seems to be an alternative since it would give control over data to the users and would enable them to work with their data in a more structured way.
This would be helpful for multiple reasons:
1) You would not loose insights from users, since they would be kept in a somewhat ordered fashion
2) Although probably a nightmare for any DBA, I could imagine it is still way better than spreadsheets, since
you can at least enforce some bounds (like MS Access for example)
3) Users could query it and use any frontend they want with a simple SQL connection (no fiddling with proprietary connectors etc.)
What are drawbacks of this? Has this already been done/tried outside of academia?
But especially for heavily regulated or understaffed companies, where development cycles are long, this seems to be an alternative since it would give control over data to the users and would enable them to work with their data in a more structured way.
This would be helpful for multiple reasons: 1) You would not loose insights from users, since they would be kept in a somewhat ordered fashion 2) Although probably a nightmare for any DBA, I could imagine it is still way better than spreadsheets, since you can at least enforce some bounds (like MS Access for example) 3) Users could query it and use any frontend they want with a simple SQL connection (no fiddling with proprietary connectors etc.)
What are drawbacks of this? Has this already been done/tried outside of academia?