Greenplum Architecture and Data Modeling

• An understanding of relational databases and SQL
• Experience with data warehousing or analytical systems
• Familiarity with Linux command line operations

Audience

• Data architects and engineers
• Database administrators and technical leads
• BI developers and analytics specialists working with Greenplum

Greenplum is an open-source massively parallel processing (MPP) data platform designed for analytics and large-scale data environments.

This instructor-led, live training (online or onsite) is aimed at intermediate-level to advanced-level technical users who wish to understand the internal architecture of Greenplum, build optimized data models, and apply high-performance SQL in distributed environments.

By the end of this training, participants will be able to:

• Understand Greenplum’s shared-nothing architecture and parallelism model.
• Design distributed table structures and partitions optimized for query performance.
• Use indexing strategies, storage formats, and physical design techniques.
• Interpret EXPLAIN plans and optimize queries using best practices.

Format of the Course

• Interactive lecture and discussion.
• Lots of exercises and practice.
• Hands-on implementation in a live-lab environment.

Course Customization Options

• To request a customized training for this course, please contact us to arrange.

Greenplum Architecture

• Parallel processing and symmetric multi-processing
• Segment roles and cluster configuration
• Scalability and data movement
• Greenplum Data Warehouse architecture

Greenplum Table Structures

• Distributed vs. randomly assigned tables
• Heap vs. append-only tables
• Row vs. columnar storage formats
• Partitioned and clustered tables

Data Distribution and Hashing

• Hashing logic and distribution keys
• Skew handling and performance impact
• Hash maps and row placement strategies

Indexes and Performance Optimization

• Clustered and non-clustered indexes
• B-tree and bitmap index use cases
• Index scan and storage behavior

Physical Database Design

• Normalization and logical model design
• User access strategies and distribution analysis
• Data demographics and indexing decisions

Denormalization Techniques

• Derived data, summary tables, and pre-joins
• Columnar tables as vertical partitioning
• Data marts and materialized views

Advanced SQL and Query Execution

• Join strategies and redistribution
• OLAP and window functions
• Temporary tables, subqueries, and derived tables

EXPLAIN Plans and Query Tuning

• Reading and interpreting EXPLAIN output
• Cost analysis and plan optimization
• Join movement and segment-local operations

Greenplum Utilities and Best Practices

• ANALYZE and VACUUM
• Data loading and movement with Nexus
• Security, permissions, and performance tips

Summary and Next Steps