Snowflake concepts & architecture
summary
Snowflake is a fully managed, cloud data platform that runs on Azure, AWS or GCP and can replicate across them via cloud-agnostic Snowgrid. Storage is decoupled from compute: query engines run in virtual warehouses (MPP compute) that auto-scale and auto-pause, while data lives in optimized compressed columnar stores. Enterprise buys 90-day time travel for point‑in‑time recovery and the engine offers zero-copy cloning so 5 TB table copies don’t duplicate storage. Demos use Azure and a 30‑day $400 free trial.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:11Snowflake as cloud platform
- 5:47Architecture layers
- 9:22Virtual warehouses cost
- 13:13Web portal features
- 17:01Cloning and time travel
- 19:50Security roles
- 20:10Summary
related talks
23mAcceleratinag Data Science and Machine Learning with Snowflake DataTech
13mLightning Talk: Highlights of the Snowflake Elastic Data Warehouse
33mData Sharing for State and Local Government
20mBuilding An Elastic Query Engine on Disaggregated Storage
15mBeyond the Silo: Modernizing SAP with Snowflake on AWS for AI
22mEmerging Models for Data Applications
26mFrom Data Sharing To Data Collaboration
27mNeo4j GDS for Snowflake