Databases for Beginners
by Sergey Okinchuk
Graph Databases (and the Family Photo)Topic 48

Chapter Eleven · Choosing

Choosing and What's Next

Ten chapters of tables, queries, guarantees, and families converge on one working skill: given a project, choose its database with reasons attached. This chapter hands you the four-question framework and drills it on real scenarios, gives honest portraits of the five engines you'll actually hear named, and explains the OLTP/OLAP split that accounts for a whole second industry. Then the last page tells you plainly what you can do now, what was left out on purpose, and where each missing piece lives.

4 topics

Back in Chapter 1, Lora chose a database without knowing she was choosing: the Marquee's data was tables-shaped, the questions crossed those tables freely, and seat G7 could never be sold twice — so relational fit, and everything after followed from that fit. This chapter turns her accident into your method. The choice becomes four questions you can ask about any project, plus one prior that has quietly run the whole book: relational by default, something else when a specific answer says so.

Around the framework sit three pieces of context every working conversation assumes. The engine names — PostgreSQL, MySQL, SQLite, SQL Server, Oracle — get one honest paragraph each, because job postings speak in those names and the differences matter less than the folklore claims. The split between databases that run the business and databases that analyze it gets its own page, because "should this query run on the till?" is a question the reader can now answer. And the final page closes the book the way it opened: honestly, with a map.

The choosing framework — four questions, one default
Is the data nested and varying, fighting fixed columns?→Document family
Is the only question a key lookup, at enormous speed?→Key-value store
Do the questions walk connections many hops deep?→Graph database
Must writes flood across many machines by design?→Wide-column store
Tables, cross-cutting questions, rules that must hold?→Relational
Not sure yet?→Relational (default)

Topics in This Chapter

Topic 48
How to Choose a Database
The four questions — shape, questions, guarantees, scale — armed with everything the book taught, the relational default defended one last time, and three project drills walked through with reasons attached.
Choosing
Topic 49
The Engines You'll Hear About
PostgreSQL, MySQL, SQLite, SQL Server, Oracle — honest one-paragraph portraits, not rankings, plus the accent strip that shows how little the choice changes for everything this book taught.
Engines
Topic 50
OLTP vs OLAP: Databases for Questions
Booking seat G7 and analyzing three years of revenue are opposite workloads. The split that explains warehouses, BigQuery, and the entire analytics industry — in one honest page.
Concept
Topic 51
Where to Go From Here
What you can actually do now, what was left out on purpose and where each piece lives, the one habit that turns reading into fluency — and the last query in the book, printed untouched.
Concept