Sequences are a feature that some DBMS products implement to provide users with a mechanism to generate unique values - the Sequence ensures that each call to it returns a unique value. This is particularly important when the Sequence's result is used as a Primary Key. These can be generated with a schema for loading onto the DBMS server.
Sequences are provided so that database users are not forced to implement their own unique value generator. Not all DBMS products support Sequences; those that do not instead provide functionality for columns to be initialized with an incrementing value.
In Enterprise Architect, Sequences can be modeled in one of two ways:
|·||As individual objects (the default method) or|
|·||As Operations in a container|
Functionally the two methods result in the same DDL being produced. The main difference is visual - by having several Operations in one container, you have fewer elements and less clutter on the diagram.
Sequences modeled as individual objects are UML Classes with the stereotype «dbsequence»; you create these by dragging the Sequence icon onto a diagram from the Data Modeling Toolbox.
Operations in a Container
Sequences modeled as operations have a container object, this being a UML Class with the stereotype «sequences» (with an s on the end). Each Sequence is an operation with the stereotype «sequ». The system provides a dedicated maintenance window through which the modeler can easily manage the Sequences defined as operations.