More Database Design Mistakes To Avoid
Duplicate data. Designed into the system. How can I count the ways?
Duplicate rows:
Perhaps caused by not using primary keys. Or, by using surrogate keys, but not putting a unique index on the natural key.
Duplicate fields:
Perhaps from denormalizing.
Duplicate tables:
Perhaps from replication. The table from the OLTP system is duplicated in the DWH. Staging tables. Or from splitting a table vertically.
Duplicate schemas:
For development and testing.
Duplicate databases:
Perhaps from a failover system.
And combinations, thereof.
Some of these may be very well thought out, and actually deliberate.
But so often they are not. The dupe data just happens.
If the duplicated data does not match, which values are correct?
This is just a quick muse. What did I miss?