Breaking Down Data Warehouses: What to Know Before You Choose
summary
Defines data warehouse by concrete tradeoffs: columnar vs row storage shows a ~1 GB sample (100 projects × 3,000 issues × 5 years × ~700 bytes) reads in ~8s row versus ~2.3s columnar (~3–4× faster); contrasts SMP versus MPP, OLTP/OLAP/HTAP, and developer productivity. Shows actual tools and features (Redshift Snowflake Postgres), demos create table as select transforms, and sketches cost math (Snowflake credits $2 per hour, 730h/month, $23 per TB). Highlights sharing, masking, and where hidden labor appears.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 2:53Why organizations need a data warehouse
- 10:45Building a data warehouse
- 16:28Access patterns and modeling
- 22:42Row vs column storage
- 26:11Key decisions before building
- 31:53Cost factors and vendors
- 36:48Q&A
related talks
58mUnderstanding Big Data for Software Engineers
32mHow open source is re-shaping the cloud data warehouse landscape
40mTo Warehouse or to Lakehouse? And How Do Various Open Source Project Compare?
31mData Warehouses Meet Data Lakes
42mBuilding Tomorrow's Warehouse Today: Modern Data Warehouse Modelling Insights
27mBuilding a PostgreSQL data warehouse
23mThe Best Data Warehouse is a Lakehouse
01h 01mModern Data Warehousing Strategy