Snowflake architecture (storage, compute, services layers) - Mini Project: Build & Apply

Snowflake Architecture: Storage, Compute, and Services Layers

📖 Scenario: You are working as a cloud architect for a company that wants to understand how Snowflake organizes its cloud data platform. Snowflake separates its architecture into three main layers: storage, compute, and services. This separation helps the company scale efficiently and manage data securely.

🎯 Goal: Build a simple Snowflake setup that demonstrates the separation of storage, compute, and services layers by creating a database, a warehouse, and a role with specific privileges. This will help you see how Snowflake manages data storage, query processing, and access control separately.

📋 What You'll Learn

Create a database named company_data to represent the storage layer

Create a warehouse named compute_wh to represent the compute layer

Create a role named data_analyst to represent the services layer

Grant the data_analyst role usage on the warehouse and select privileges on the database

💡 Why This Matters

🌍 Real World

Snowflake's architecture allows companies to store large amounts of data securely, process queries efficiently, and manage user access separately. This separation helps in scaling and optimizing costs.

💼 Career

Understanding Snowflake's architecture is essential for cloud architects, data engineers, and database administrators who design and manage cloud data platforms.

1

Create the storage layer database

Create a database called company_data to represent the storage layer in Snowflake.

Snowflake

# Create the database company_data
# Your code here

Need a hint?

Use the CREATE DATABASE command followed by the database name.

2

Create the compute layer warehouse

Create a warehouse called compute_wh to represent the compute layer in Snowflake.

Snowflake

CREATE DATABASE company_data;
# Create the warehouse compute_wh
# Your code here

Need a hint?

Use CREATE WAREHOUSE with a small size and enable auto suspend and resume for cost efficiency.

3

Create the services layer role

Create a role called data_analyst to represent the services layer in Snowflake.

Snowflake

CREATE DATABASE company_data;
CREATE WAREHOUSE compute_wh WITH WAREHOUSE_SIZE = 'XSMALL' WAREHOUSE_TYPE = 'STANDARD' AUTO_SUSPEND = 60 AUTO_RESUME = TRUE;
# Create the role data_analyst
# Your code here

Need a hint?

Use CREATE ROLE followed by the role name.

4

Grant privileges to the role

Grant the data_analyst role usage on the compute_wh warehouse and select privileges on the company_data database.

Snowflake

CREATE DATABASE company_data;
CREATE WAREHOUSE compute_wh WITH WAREHOUSE_SIZE = 'XSMALL' WAREHOUSE_TYPE = 'STANDARD' AUTO_SUSPEND = 60 AUTO_RESUME = TRUE;
CREATE ROLE data_analyst;
# Grant privileges to data_analyst role
# Your code here

Need a hint?

Use GRANT USAGE ON WAREHOUSE and GRANT SELECT ON DATABASE commands for the role.