Assign entities to areas
You need to assign entities to database areas to provide for efficient application runtime processing:
By placing Commonweather entities in separate areas, we enable programs to prepare only the area or areas required for a particular operation rather than the entire database. In addition, we reduce the likelihood of contention for heavily-used entities.
You might want to assign entities and indexes to separate areas.
Compress entities
The JOB and INSURANCE PLAN entities each contain a data element that provides descriptive information about a particular entity occurrence (JOB DESCRIPTION and PLAN DESCRIPTION). As such, these entities are good candidates for compression.
Relationship options
All relationships are linked to provide most efficient access. Since this is not a large database, it is not necessary to eliminate relationships between areas.
All clustered relationships are chained; all nonclustered relationships are indexed. This reduces I/O when accessing nonclustered relationships and reduces CPU when accessing clustered relationships.
The following relationships are sorted with the unique option to eliminate indexes used only to enforce unique constraints:
New Sorted Relationship |
Sort Key |
Index Eliminated |
---|---|---|
EMP-EMPOSITION |
START DATE |
JOB-NDX |
EMP-EXPERTISE |
SKILL CODE |
EMP-NDX |
NHC-PROC |
PROCEDURE NUMBER |
NON-HOSP-NDX |
DC-PROC |
PROCEDURE NUMBER |
PROC-NDX |
The following relationships are sorted to avoid sorting retrieval occurrences:
Sorted Relationship |
Sort Key |
---|---|
DEPT-EMPLOYEE |
EMP LAST NAME EMP FIRST NAME |
OFFICE-EMPLOYEE |
EMP LAST NAME EMP FIRST NAME |
SKILL-EXPERTISE |
SKILL LEVEL |
All sorted relationships are order ascending except:
Copyright © 2014 CA.
All rights reserved.
|
|