Entry Points

Consider the following rules:

Names of entry points must be unique within a model.
The name of the entry point must be the same as that of the data store indexspace.
A record can have only one primary key.
A record can have only one clustered entry point.
A record can have only one partitioned entry point.
If an entry point implements a many-to-many relationship implementation, then the entry point must be on a link record.
An entry point that implements an identifier must match the identifier.
If the entry point was added for performance reasons, then the identifier ID and the relationship ID fields should be blank.
For an entry point to be partitioned, the data store indexspace must contain multiple data sets.
For a non-partitioned entry point, the data store indexspace cannot contain multiple data sets.

The following table describes the content of each field type:

Field Name	Content
MODEL_NAME	Host Encyclopedia name of the model
RECORD_TYPE	O (Object Record)
OBJECT_TYPE	ENTRY_POINT
OBJECT_NAME	Name of the entry point
REF_OBJ_1	Name of the data store indexspace that stores this entry point
REF_OBJ_2	Name of the record referenced by this entry point
REF_OBJ_3	Unique ID of identifier, if entry point implements an identifier
NUM_PROP_1	Unused
NUM_PROP_2	Unused
NUM_PROP_3	Unused
NUM_PROP_4	Unused
NUM_PROP_5	Unused
NUM_PROP_6	Unused
NUM_PROP_7	Unused
NUM_PROP_8	Unused
CHAR_PROP_1	Is the entry point clustered? Y (Yes) N (No)
CHAR_PROP_2	Is this entry point unique? Y (Yes) N (No)
CHAR_PROP_3	Is this the primary key for the record? Y (Yes) N (No)
CHAR_PROP_4	Is this entry point partitioned? Y (Yes) N (No)
CHAR_PROP_5	Unused
CHAR_PROP_6	Unused
CHAR_PROP_7	Unused
CHAR_PROP_8	Unused
NAME_1	Unique ID of the relationship implementation (entry point or many-to-many), if the entry point is associated with a relationship implementation
NAME_2	Unused