O3 Builder is the component responsible for building the cube files from the models defined in O3 Designer.

Using the O3 model definition from the .mdl file generated in O3 Designer, you can retrieve data the data from the specified sources and build the cube for use with O3 Browser.
You can build cubes using one of the following methods:

You can also schedule cube building through the operating system's task scheduler. This solution is ideal for those cases where building of the cube, due to its size, takes a long time and impacts on resources. This mode allows scheduling of the cube building for a time when resource consumption is low.
You should also use the task scheduler for updating existing cubes on a regular and automatic basis. Using the task scheduler, you can program routine cube rebuilding tasks for specific time intervals such as daily or weekly.
O3 Builder must have access to, from the machine it executes on:

• text file directory locations for text data sources
• SQL data sources through JDBC/ODBC connections

O3 Builder can send information about the final status of builds through e-mail. In the model definition, you can determine e-mail recipients based on the build type (incremental or complete) and the build result (OK or errors). Using the model definition, you can also determine who should be informed in different situations such as: complete updating OK, incremental updating OK, or errors.
You can also specify that in the case of certain errors, such as when the database source for retrieving data is inaccessible, the build should retry a given number of times after a given interval.

Full Builds

When you execute the build from O3 Designer, the generated cube corresponds to the model currently active in O3 Designer.
If you wish, you can change the order that elements of each dimension hierarchy's level displays in for the O3 Browser prior to a complete build.
Notes

• If you change the order prior to an incremental build, the changes are ignored.

Incremental Builds

Use an incremental cube build to update cube information. This avoids complete rebuilds of cubes with lengthy build times.
Two features support the incremental build:

Data sources forming the incremental input set need only provide the cube's new data. If these data sources retrieve already included data, the cube will contain erroneous data. When defining the data sources:

• ensure that the text data source contains only records to be added
• limit SQL query data sources by using appropriate WHERE expressions

For SQL queries, you can include build parameters in the WHERE expression that limit the records obtained in accordance with the value they take in each build. The WHERE expressions can contain conditions dependent on record dates or ranges of values such as invoice numbers.
We recommend conditions based on dates as their management is easily automated through build parameters, as shown in the following example.