DataStage - Intermediate

    

Summary:
The Intermediate course is designed to reinforce the Parallel nature of the datastage environment. A quick review of the parallel nature of the environment will be undertaken. More advanced concepts such as optimization of job design by using Multi-stage take-on of data. Breaking up job execution to take advantage of the parallel nature of the Enterprise engine. Examples such as performing both an aggregation of data and joining of datasets will be explored. Understand how the execution of jobs may be impacted by your APT Configuration file and fin-tuning of the APT file.

Duration:
3 Days/Lecture & Lab

Audience:
This class is designed for designers in the Datastage Enterprise environment.

Topics:

• Overview of ETL, DataStage and software available, DataStage Enterprise Edition tools (Admin, Designer, Director, and Web Administrator)
• Parallel Concepts (Application scalability, Explicit & Implicit parallelisms, Pipeline and partition parallelisms, Framework of engine)
• Datasets and their types (Virtual -links between stages & Persistent - physical datasets)
• Partitions and Collectors (review of types including hash and random and stages such as aggregator, join and merge which are dependent on partitioners)
• Importing and Exporting of data (how to connect metadata using plugins and ODBC)
• Overview of Enterprise Edition stages which include Transformer, Filter, Switch and Aggregator)
• Complex connection with Oracle, how to use an Oracle grid with DataStage, how Upsert vs. Design update and insert
• More Enterprise Edition stages and how to improve performance such as Row to Column conversion
• Create Subrecord.

Prerequisites:
Students should have a minimum of 6 months development experience in the DataStage environment.

Last Update: May 20, 2012