Oracle® Business Intelligence Data Warehouse Administration Console User's Guide Version 10.1.3.4 Part Number E12652-02 |
|
|
View PDF |
This chapter describes basic concepts pertaining the Oracle Business Intelligence Data Warehouse Administration Console.
This chapter contains the following topics:
DAC provides a framework for the entire life cycle of data warehouse implementations. It enables you to create, configure, execute, and monitor modular data warehouse applications in a parallel, high-performing environment. For information about the DAC process life cycle, see "About the DAC Process Life Cycle".
DAC complements the Informatica ETL platform. It provides application-specific capabilities that are not prebuilt into ETL platforms. For example, ETL platforms are not aware of the semantics of the subject areas being populated in the data warehouse nor the method in which they are populated. DAC provides the following application capabilities at a layer of abstraction above the ETL execution platform:
Dynamic generation of subject areas and execution plans
Dynamic settings for parallelism and load balancing
Intelligent task queue engine based on user- defined and computed scores
Automatic full and incremental mode aware
Index management for ETL and query performance
Embedded high performance Siebel OLTP change capture techniques
Ability to restart at any point of failure
Phase-based analysis tools for isolating ETL bottlenecks
Important DAC features enable you to do the following:
Minimize installation, setup, and configuration time
Create a physical data model in the data warehouse
Set language, currency, and other settings
Design subject areas and build execution plans
Manage metadata driven dependencies and relationships
Generate custom ETL execution plans
Automate change capture for the Siebel transactional database
Capture deleted records
Assist in index management
Perform dry runs and test runs of execution plans
Provide reporting and monitoring to isolate bottlenecks
Perform error monitoring and email alerting
Perform structured ETL analysis and reporting
Utilize performance execution techniques
Automate full and incremental mode optimization rules
Set the level of Informatica session concurrency
Load balance across multiple Informatica servers
Restart from point of failure
Queue execution tasks for performance (See Figure 3-1.)
DAC manages the task execution queue based on metadata driven priorities and scores computed at runtime. This combination allows for flexible and optimized execution. Tasks are dynamically assigned a priority based on their number of dependents, number of sources, and average duration.
DAC is used by different user groups to design, execute, monitor, and diagnose execution plans. These phases together make up the DAC process life cycle, as shown in Figure 3-2.
The phases of the process and the actions associated with them are as follows:
Setup
Set up database connections
Set up ETL processes (Informatica)
Set up email recipients
Design
Define application objects
Design execution plans
Execute
Define scheduling parameters to run execution plans
Access runtime controls to restart or stop currently running schedules
Monitor
Monitor runtime execution of data warehouse applications
Monitor users, DAC repository, and application maintenance jobs
Source system containers hold repository objects that correspond to a specific source system. For information about the different kinds of repository objects, see "About DAC Repository Objects Held in Source System Containers".
You should use the preconfigured source system containers to create your own source system container. The DAC client lets you do modifications to objects in your own source system container built as a copy of the preconfigured containers. This enables the DAC client to track the customizations, such as newly created objects, modified objects, and those that are used as is.
Caution:
You should not modify objects in the preconfigured source system containers either through the DAC client or directly through SQL statements to the DAC repository. You must make a copy of a preconfigured container in order to make any changes to it.
For instructions on creating a new source system container or copying an existing container, see "Creating or Copying a Source System Container".
All DAC repository objects are associated with a source system container. For more information about source system containers, see "About Source System Containers" and "About Object Ownership in DAC".
DAC repository stores application objects in a hierarchical framework that defines a data warehouse application. DAC enables you to view the repository application objects based on the source system container you specify. The source system container holds the metadata that corresponds to the source system with which you are working.
A data warehouse application includes but is not limited to the following repository objects:
Subject areas. A logical grouping of tables related to a particular subject or application context. It also includes the tasks that are associated with the tables, as well as the tasks required to load the tables. Subject areas are assigned to execution plans, which can be scheduled for full or incremental loads.
Tables. Physical database tables defined in the database schema. Can be transactional database tables or data warehouse tables. Table types can be fact, dimension, hierarchy, aggregate, and so on, as well as flat files that can be sources or targets.
Indexes. Physical database indexes to be defined in the database schema either to better the performance of the ETL processes or the queries for reporting purposes.
Tasks. A unit of work for loading one or more tables. A task comprises the following: source and target tables, phase, execution type, truncate properties, and commands for full or incremental loads. When you assemble a subject area, DAC automatically assigns tasks to it. Tasks that are automatically assigned to the subject area by DAC are indicated by the Autogenerated flag in the Tasks subtab of the Subject Areas tab.
Task groups. A group of tasks that you define because you want to impose a specific order of execution. A task group is considered to be a "special task."
Execution plans. A data transformation plan defined on subject areas that needs to be transformed at certain frequencies of time. An execution plan is defined based on business requirements for when the data warehouse needs to be loaded. An execution plan comprises the following: ordered tasks, indexes, tags, parameters, source system folders, and phases.
Schedules. A schedule specifies when and how often an execution plan runs. An execution plan can be scheduled for different frequencies or recurrences by defining multiple schedules.
The source system container in which an object originates is the owner container. The tabs in the DAC Design view display the owner of the various repository objects. You can reuse an object among different source system containers by referencing the object. A reference works like a symbolic link or shortcut. You can use the referenced object just as you would an original object, but the object's ownership remains unchanged.
For example, W_INVOICE_F is a fact table whose owner is the data warehouse source system container. You can reuse W_INVOICE_F in any other container by referencing it.
You can reference an object from its owner container, and you can also reference an object that has already been referenced by another source system container.
If you modify a referenced object, the modified object becomes a clone and the ownership changes to the source system container in which you performed the modification.
When you make changes to an original object that has been referenced by other containers, any updates to the original object are immediately reflected in the referenced object. If you delete the original object, all referenced objects are also deleted.
Changes to an original object's child objects are not automatically reflected in the referenced object's child objects. Use the right-click command and select Ownership, then select Push to References to push the changes to the referenced object's child objects. And, conversely, you can import into a referenced object the changes made to an original object; this function is referred to as a re-reference.
For a description of the ownership functionality available in the Design view right-click menu, see Table 6-6, "Design View Right-Click Menu Commands".
The DAC Import/Export feature enables you to import or export source system-specific DAC metadata into or out of the DAC repository.
Caution:
When you export DAC metadata, DAC erases all of the files in the target folder.
DAC export behavior is as follows:
If the target folder is empty, DAC exports without a warning.
If the target folder contains DAC metadata, DAC exports after warning and when OK is clicked. The process replaces all content in the target folder with a new export.
If the target folder has non-DAC metadata along with DAC Metadata, DAC exports after warning and when OK is clicked. The process replaces all content in the target folder with new export. All non-DAC metadata is deleted.
If the target folder has only non-DAC metadata, DAC cannot export into that target folder.