I'm building a database scheme for two different headquarters. Each location uses the same data-scheme, but has its own data. Both will have access to the same location, a shared network folder where the corresponding access-backend will be placed. Both will also use an equal Front-End, where depending on the user(-account) the specific data will be shown. Each location will have two main tables which will grow by ~4000-5000 records per year.
Most of the time each location only uses its own data, but once every week or so a third party will access some aggregated data from both locations.
In my opinion there are two possibilities to organize the data: a shared approach, where both locations use & work with the same tables, where a field indicates whether the current data belongs to location 1 or 2. A huge problem I see is the amount of data which would be transferred through the network (roughly doubled), because (as far as I know) access runs the select statements locally, after transferring the data. (I stand corrected). Besides that an additional query would be neccessary for each access.
The alternative would be to create a second set of tables for the second location and keep them 'separated' (e.g. renaming them to loc1_tablex and/or create a second backend-file), which also would make a backup easy.
I would prefer the second approach, because I can't really see any downside to it. It would be faster and in my opinion cleaner.
Did I miss anything importing or would this approach be reasonable?
Edit:
After discussing the topic further (we also got some new information from our client), we decided to store the data in one backend and in the same tables. The table structure is fully equal and we are using a location id to control what data belongs to which location (first approach). Ty for shedding some light on the access backend, I had a lack of information there and learned a lot!