SQL - Table

List

See SQL - Schema (Metadata)

Creation

• Sql:
  • SQL - CREATE TABLE - create the structure
  • SQL - VALUES - create a table from values

Cardinality

The cardinality of a table is the size of a table, the name comes from the set theory.

Split

Even if it's possible to get all data in one table that has the same grain (unique key), you may made two tables to split the columns that are not functional related.

Why ?

• because you get minder dependency. You can changes your table and your process independently. For instance, the test data set don't need to be updated to reflect the changes of a second functional process.
• You can then set columns NULLABLE. If a process updates columns after an other one (such as total calculation), the value of this column are unknown for the first process and the columns must be set not nullable.
• You can add audit columns by process (Update and create (date|user))

Why not ?

• You may got consistency problems. There may be not a one-to-one relationship. But as the two functional processes are independent, it's may be not always needed.
• Performance: You add a join cost