Duration: Variable (1 to 5 days)

Objectives:

Having completed this class the student will be able to:

• configure partitioning, sorting and buffering optimally
• make use of inbuilt reporting and troubleshooting tools
• create jobs that work with vectors and subrecord types
• implement change detection and surrogate key management
• employ parallelism in Enterprise stage types

Contents:

The Advanced Parallel Job Techniques Class is constructed from the following thirteen modules. The class can be delivered in variable amounts of time from 1 day to 5 days. The first four modules must be included, and require one day. The remaining nine modules, if taken, must be taken in the order specified and require one half-day each. All thirteen modules can be delivered in five teaching days.

• Review of the parallel execution framework (architecture)
• Partitioning and collecting techniques
• Sorting techniques
• Buffering
• Managing resource allocation and performance tuning
• DB2 Enterprise stage
• Oracle Enterprise stage
• Teradata stages
• Monitoring: the OSH and the Score
• Vectors and Subrecords and handling structured and raw data
• Change detection techniques (vectors and subrecords module must also be selected)
• Surrogate key management in a parallel environment
• Troubleshooting and debugging