Posted - 11/24/2012 : 10:37:08
| A vendor of ours has structured the data so that all data entities are grouped within a single entity for all relations, stating that this may provide performance benefits when using SQL Server. I have never seen such a weird way of going about normalization of data, however I might not be experienced in "avante-garde" designs.
DBID PersonID PersonName PersonAddress
DBID ClientID ClientName ClientAddress ClientTelephone
DBID represents a company database. So logically each DBID refers to a separate database, whilst physically each entity would have its own DBID so that the SQL developer would be able to query each database separately. This might create some duplication of data since the DBID has to be stored as a field in each row of data.
I am not a certified SQL Server administrator, however, from a conceptual point of view I saw the design as logic which does not make any sense to me, however, I would like your opinion about it.