CSE Roadmap

Common tier · about 3 hours · 7 parts

Part 1 of 7 · 1 min read

Why this page comes after storage engines

In this module you will

  • Tell an online transaction workload (OLTP) from an analytic workload (OLAP) by their read and write patterns.
  • Explain why companies copy data into a separate data warehouse, and describe the extract, transform and load (ETL) steps.
  • Draw a star schema with a fact table and dimension tables for a simple business.
  • Compare row storage and column storage with byte counts for the same query.
  • Work through a bitmap index and run-length encoding on a short column.
  • Say how a column store takes writes, and what a materialized view saves.

The last two pages were about finding single records fast. A business also asks questions of a different kind: totals, averages and trends over millions of rows. These need a different layout on disk. Many interviews include a metrics or reporting part in the design, so this page gives the words and the reasons. The tier is Common. The numbers on this page are made up to show the idea, unless the text names a book.