You start a load process by using a project in BIRT Analytics Loader. Before running a project, you must create the data source connections and define the loading transactions and the data transformation scripts. To make the loaded data available to the BIRT Analytics users, you must synchronize the loaded database with the active database and grant access privileges to database users.
After defining the table mappings and all transactions in a load process, you can run the project to load the data. You can run a single or multiple transaction processes. You can run a process with or without data transformation.
The load process loads the data in the FastDB database, the same database that users analyze. To avoid inaccurate analysis, consider loading data when users are not logged in to the system.
To prevent user access to databases while a load process runs, stop the web server running the BIRT Analytics application. Alternatively, schedule the loading process to run at a time when users are off line. For more information on how to schedule a load process, see Administering BIRT Analytics Loader.
Understanding data loading tasks
The following tasks are part of a typical load process to load data into the FastDB engine:
Define a data loading project to collect, organize, and load data into FastDB.
Run a data loading project.
Remove cached data objects from FastDB. You can run this as a Loader—Post loading action or in BIRT Analytics Administrator.
Synchronize the FastDB repository with the ApplicationWS backend database (also known as wpt). You can run this as a Loader—Post loading action or in BIRT Analytics Administrator.
Set access permissions for data objects in the FastDB repository. You can set permissions using the Administration—Security tool.
Skipping one of these steps can cause users of the BIRT Analytics user interface to see an empty Data tree.
Running a project
To run a loading project, open the Data tab and make any desired changes. Then choose Run, as shown in . Use the other feature tabs to add data transformations and post loading actions to the loading project.
Figure 2‑8 Reviewing the Data feature
In the message that confirms whether you want to run the project, choose Yes. A list of database table names and the records loaded into each table appears in Running project, as shown in Figure 2‑9.
Figure 2‑9 Running project log
The Running project log provides information about the tables into which the data is loaded, and the number of the records that are loaded. In the log viewer, standard message text is in black print, INFO messages in green and Error messages in red. The stack trace shows the line of script where the error was generated.
To limit the number of message lines in the log, use Options—Notify. Set notify to zero to record a log message at the end of a load process for each table after all rows for the table finish loading. Set the notify value to record a log message every time the number of loaded rows equals the notify value. For example, with notify set to 100000, a log message is generated every time 100000 rows are loaded. In this case, a data resource with 1000000 rows of data generates 10 messages in the log.
For loading processes requiring a long time, choose the option Run in background. This option minimizes the Running project window and enables you to work on a different project until the running project finishes.
Using reports
To browse through the generated log files, select Reports from the control menu. The list of reports appears in Executions, as shown in Figure 2‑10. Date shows a date and time when a project is executed. State shows the generation status. After selecting a date, the log content appears in Detail.
Define a data loading project to collect, organize, and load data into FastDB.
