Imagine a supermarket with dozens of separate stores. Each store has more than twenty thousand items selling on its shelves. And each one needs to keep a close check on its sales and stock levels.
Question: does is make sense to have one gigantic database to hold every single item in every single store? Answer: No
The reason is that each store has no particular reason to know the exact sales of another store. Even the centre headquarters does not need to know the minutae of every sale - they are only interested in daily summary reports.
This is where a 'partitioned database' comes in. A partitioned database is an example of a distributed database. The exact records are held in separate database held locally, but there may be a master index to each and every record in the combined databases at the centre.
A partitioned database solves the problem of handling the details locally and yet keep the rest of the organisation informed at the appropriate level of detail.
It saves on wasted network bandwidth and it keeps information at the most efficient level for the organisation as a whole.
Challenge see if you can find out one extra fact on this topic that we haven't already told you
Click on this link: Partitioned database