ODI Learning Note 2--odi product architecture

ODI Learning Note 2--odi product architecture

ODI Product Architecture:

ODI provides the following management tools:
Designer is used to define data transformation logic, which is the most common development tool, most of the development tasks, including the definition of data store, interface (mapping relationship) and the creation of the package (equivalent to Workflow), etc. is done in the designer.
Operator is used to manage and monitor the execution of data conversion tasks, and can also be used for debugging during the design phase (debugging)
Topology Manager is used to define physical and logical infrastructures, such as creation and management of work repository.
Security Manager for managing user permissions
and a program agent:
Schedule Agent Scheduler, which is used to schedule data transformation tasks. The scheduling agent also comes with a data transformation engine, but ODI uses the E-LT architecture, so basically the planning agent simply passes the task to the target library, and its data conversion engine is seldom used.

Extensible knowledge Model (knowledge Modules)

ODI can load and transform data from a variety of heterogeneous database platforms and information queue-based technologies (WEB services), with flexibility, extensibility and effectiveness, mainly relying on knowledge model;

The knowledge model, as a plug-in for ODI, encapsulates various ETL processes for specific data sources and target. ODI provides six knowledge models, as shown in:

RKM: The function of the data model is completed by reverse engineering of the structure of the source and target systems.

JKM: Complete the data model features that capture CDC, perform incremental extraction, and work task scheduling.

LKM:LKM completes loading from the source database data into the staging table.

CKM:CKM Complete the data quality check.
IKM:IKM completes the loading of data from the staging table into the target table.
SKM:SKM completes the functionality of the ODI and Web services interfaces.

Data-quality Firewalls

By using the CKM knowledge model, only those data that conforms to the business rules are allowed to enter the integration process, which can block the data from the data source, thus guaranteeing the quality of the data!

Support for Changed-data Capture

The JKM knowledge module monitors data from source database changes, and some database vendors, such as Oracle, directly support CDC, while others capture DML statements through, for example, trigger technology.

Oracle Data Integrator in Relation to Oracle Warehouse Builder

OWB is powerless in terms of SOA and non-Oracle data sources, but the two are complementary relationships, see:

ODI can be derived from a wide variety of heterogeneous data sources, including Web services and event-based architectures. Once the data is integrated into the DW Staging AREA,OWB begins taking over the next tasks, including creating ODS and multidimensional models.

See:

Http://www.oracle.com/technology/pub/articles/rittman-odi.html

Http://snipkingderek.spaces.live.com/blog/cns!EE24905211C3FFD!399.entry

ODI Learning Note 2--odi product architecture