HP Intelligent Management Center Standard Software Platform - Manual (Page 667)

When a one-to-many situation exists between two database tables and the
program matches up records from the tables, there are a number of factors
that determine how many records the program reads and evaluates.
The tables that follow show the effects of the different factors on the number
of records the program ultimately has to read. The charts are based on these
assumptions:
• Table A contains 26 records (one for each letter in the alphabet).
• Table B contains 2600 records (100 matching records for every record
in Table A).
• The scenario is to produce a report that finds two specific records in Table
A and the 200 records (100+100) in Table B that match those two records
in Table A.
In a best case scenario, the program would only have to read about 200
records to accomplish the task.
In a worst case scenario the program would have to read about 67,600
records to accomplish the same task.
Note:
The performance considerations for data files are different from the
considerations for SQL databases. A data file is any non-SQL database that
is accessed directly from Crystal Reports. For the purpose of this discussion,
an SQL database is any database capable of accepting SQL commands
accessed directly from Crystal Reports or through ODBC as well as any
other database types that are accessed through ODBC.
For a better understanding of the difference between direct access databases
and ODBC data sources, see Accessing Data Sources in the online help.
Extended descriptions of chart columns
The performance charts use the following columns:
• Linking or Subreport
Are you creating a report from linked databases or are you inserting a
subreport and binding it to the data in your primary report?
• Selection Formula
Crystal Reports 2008 SP3 User's Guide 667
25 Understanding Databases
Linking tables