Query Performance in Snowflake

Introduction to Data Modeling in Snowflake

Nuno Rocha

Director of Engineering

Data storage

Row-based storage

Data storage (1)

Row-based PostgreSQL

Data storage (2)

Row-based Snowflake

Massively parallel processing

Multiple parallel processes executing a query

MPP's role in data storage

Data split in micro partitions

MPP's role in data storage (1)

Data split in regular partitions

Visualizing query execution times

Query execution result

Visualizing query execution times (1)

Instruction to find query profile

Visualizing query execution times (2)

Query profile page

Terminology overview

Massively Parallel Processing (MPP): Snowflake's engine that processes data using multiple servers simultaneously.
Micro-partitions: Small data storage segments in Snowflake that enhance retrieval speed.
Columnar Storage: Data storage format that stores each column of data separately.
Row-based Storage: Traditional data storage format where each row of data is stored sequentially.
PostgreSQL: Open-source relational database system that uses row-based storage, commonly compared to Snowflake for performance benchmarking.
Query Profile: Snowflake feature to visualize the steps and resource usage of query execution.

Let's practice!

Introduction to Data Modeling in Snowflake

Preparing Video For Download...