1. Trade-Offs in Data Systems Architecture
1.8 Terminology introduced here (used for the rest of the book)
Terminology introduced here
38- frontend / backend
- data infrastructure
- stateless
- data-intensive
- OLTP
- point query
- OLAP
- BI
- data warehouse
- ETL / ELT
- data pipeline
- reverse ETL
- HTAP
- data lake
- sushi principle
- system of record / source of truth
- derived data
- cloud native
- IaaS
- on premises
- object storage
- separation of storage and compute
- disaggregation
- multitenancy
- DevOps / SRE
- metered billing
- distributed system
- node
- elasticity
- data residency
- observability
- SOA / microservices
- serverless / FaaS
- HPC
- bisection bandwidth
- Clos topology
- data minimization (Datensparsamkeit)
- right to be forgotten
1.7 Decision cheat sheet
Yes if any of: analysts need to join across ≥2 operational systems; analytical queries would compete with user traffic; analysts need ad-hoc SQL; or dataset is heading past a few…
1.9 Worked examples
Analysts need to join across 3 of them. Data volume says "no warehouse needed" — but the silo argument says yes, because you cannot join across three separately-owned OLTP databas…