Below is the 54 points/checks one should consider for developing universe .These points can be best used at the time of unit testing of Universe.These point may insure to deliver zero defect Universe design.
No.
|
Particulars
|
Peer-review
|
Yes/No/NA
| ||
Connection
| ||
1
|
Connection to the Database is according to the requirement
| |
2
|
Parameters - SQL Tab - Multiple SQL statements for each measure is unchecked
| |
3
|
Parameters - SQL Tab - Cartesian Products - Prevent is checked
| |
4
|
Database Connection named or define in such a way that it refers to actual database
| |
Tables
| ||
5
|
Alias Tables are named with proper functional use.
| |
6
|
For Aliasing tables, the original one noted with ‘Aliased Table – Do Not Remove’
| |
7
|
Tables are arragned in the Structure as per Business/Functional logic
| |
8
|
Tables are connected or joined.
| |
9
|
Isolated tables are placed in separate universe.
| |
10
|
Context are named with proper functional use.
| |
11
|
All joins falls under some meaningful named context
| |
12
|
Cardinalities are set.
| |
Class
| ||
13
|
Universe Classes / Subclasses Defined as per the business logic & Naming Convention
| |
14
|
Description for the use of each Class/Subclass is given
| |
15
|
Maximum number of the objects in the classes are restricted to 30
| |
Objects
| ||
16
|
Objects used in calculation are of type Measure.
| |
17
|
Objects used in Analysis are of type Dimension.
| |
18
|
Object Description is given properly.
| |
19
|
LOV Option for each Dimension is set only for required Objects
| |
20
|
LOV option for each measure objects is unchecked.
| |
21
|
"Automatic Refresh before Use" option is checked for LOV Objects.
| |
22
|
Significant name is given to the list name under Object properties for editible LOV.
| |
23
|
Duplicate Object names are there in different classes.
| |
24
|
Objects and classes naming convention is as per business terms
| |
25
|
Timestamp is not present in the LOV's of properly defined Date prompt.
| |
26
|
Object formatting specified at universe level, especially with measures. (Formatting like dates, currency etc.)
| |
27
|
Multiple SQL statements for each measure should be turned off, to generate optimum SQL. (This option should be checked if the measures being retrieved in the same query involve different tables)
| |
28
|
All the Objects are arranged in alphabetical order within calss and classes are arranged ina hiearachial order from highest to lowest grain.
| |
29
|
WHERE' Clause is not used in any objects at the time of creation in universe.
| |
30
|
HTML Objects like hyperlink are created at Universe level.
| |
31
|
Only required fields of a Table are included as Objects instead of the entire fields of the Table.
| |
Predefined Conditions and Hierarchies
| ||
32
|
Condition Objects are created mostly at Universe level.
| |
33
|
Description is given for each pre-defined condition.
| |
34
|
LOV is provided for a associated dimention when condition is resulting in a prompt.
| |
35
|
Time dimension related predefined conditions are expressed as Current year, Current month, previous year, last (x) weeks etc making it easy for scheduling daily/weekly/period based reports.
| |
36
|
Custom hierarchies are defined and done according to the end-user’s requirements
| |
37
|
Regularly used conditions e.g. current year / latest date do not have prompts.
| |
Import/Export
| ||
38
|
The path for Import, is in the Business Objects' Universe folder.
| |
39
|
Universe is locked when Administrator\ Designer does not want any user to Import/Export.
| |
40
|
Before Exporting Universe 'Integrity check' is performed
| |
41
|
Folders are structured in such way that secured environment is maintained.
| |
42
|
While deleting any object from an exported Universe ,impact analysis is performed.
| |
Use Aggregates for measures
| ||
43
|
Aggregation is used in all measure objects
| |
44
|
In measures SQL aggregation functions are present.
| |
Use Partitions for high volume fact tables
| ||
46
|
To boost query performance partition of high volume fact tables is used.
| |
47
|
To improve query performance aggreate tables,materialised views with aggregate awareness are included.
| |
48
|
Arrey fetch size parameter optimal value is set up.
| |
Other Considerations
| ||
49
|
Short cut Joins are used where ever possible to improve performance.
| |
50
|
Integrity check and database straucture refresh is performed.
| |
51
|
Unnecessary Objects are removed from universe before moving it to production.
| |
52
|
Universe is created from the end-user perspective, and not from a developer’s perspective.
| |
53
|
Length of query time and Number of rows retrived is set as per the database.
| |
54
|
Universe is tested based on the functional requirement
|