CSE Roadmap

Core tier · about 4 hours · 8 parts

Part 1 of 8 · 1 min read

Why this page comes seventh

In this module you will

  • Explain why redundancy is needed and what it does not cover.
  • Choose a failure domain: machine, rack, zone, region or provider.
  • Compare backup and restore, standby and active-active designs.
  • Define recovery time and recovery point objectives and work out what each design can promise.
  • Explain why replication is not a backup, and how to prove a restore works.

Pages 2 to 6 deal with failures a running system can absorb. This page deals with the loss of the thing the system runs on: a machine, a data centre, a region, or the data itself. The answer is to have more than one of each, and a plan for the day they are all gone. The books used are "Designing Data-Intensive Applications" (DDIA), chapters 1 and 7, "System Design Interview" (Xu), chapters 1 and 6, "Understanding Distributed Systems" (UDS), chapter 14, and "Building Microservices" (BM), chapter 11. The terms recovery time and recovery point objective are not in these books. They are standard industry terms, and the text says so where they appear. The numbers are made up to show the idea, unless the text names a book.