Module 2 · Snowflake Deep Dive
The single biggest investment in this course. Snowflake is listed first in the target role's skill list and you arrive with zero Snowflake-specific experience. We treat it as a full, hands-on build-out on the free tier.
Required skill · s1 (paraphrased)
Snowflake: advanced SQL, stored procedures, warehouse architecture, storage layer, Time Travel, Zero-Copy Cloning.
Your coverage: Gap
What we're assuming
- SQL fluency: you've written SQL across multiple engines.
- Procedural programming: Python and Bash both native; procedures will be the smallest part of this module.
- Column-store warehouse mental model: if you've used a columnar MPP appliance earlier in your career, similar concepts apply — but with a very different ops model.
What we modify for your background
- Skip: SQL joins/GROUP BY/window function basics. We go straight to Snowflake-dialect specifics.
- Emphasize: separation of compute and storage (new vs on-prem MPPs), Time Travel + Zero-Copy Cloning (unique Snowflake features, called out in the skill requirement), micro-partitions & clustering (the "why" behind Snowflake's perf).
- New: virtual warehouse lifecycle, result cache, Streams + Tasks (native CDC).
Want more depth?
See Credits → Module 2. Top picks: Snowflake Quickstarts (free, hands-on), select.dev blog (cost optimization), and the docs themselves.
Hands-on setup required
This module leans heavily on a live Snowflake trial account. Exercise 2.1 walks you through signup (no credit card, $400 credits, 30 days) and the cost guardrails you set before running any query. Do that first.
Lessons
- Lesson 2.1 — Warehouses, Storage, and the Architecture That Makes Snowflake Snowflake crash deep
- Lesson 2.2 — Advanced SQL in Snowflake (dialect, QUALIFY, LATERAL, semi-structured) crash deep
- Lesson 2.3 — Stored Procedures (SQL Scripting & Python) crash deep
- Lesson 2.4 — Time Travel & Zero-Copy Cloning crash deep
- Lesson 2.5 — Streams, Tasks, and Snowflake-native CDC deep