Course
Database Internals
What really happens inside a database when you run a query, write a row, add an index, replicate or shard. Short lessons on one shop dataset, real PostgreSQL output from a lab you run yourself, and two engines you build in Go: TinyDB, a storage and query engine, and ShardKV, a replicated, sharded key-value store.
from a SQL query to a distributed database ↓
Route ticket
From
a SQL query
To
a distributed database
- Stations
- 123
- Parts
- 14
- Reading time
- 71–101 h
- Languages
- SQL (PostgreSQL) · Go
contents
The route
14 parts · 123 stations · 71–101 h
Terminal
Ride the whole line and you have built three systems: a storage engine with a B+Tree, joins and crash recovery, a sharded, replicated store that survives killed nodes, and ShopDB, a distributed SQL database that keeps money and stock correct through crashes and partitions. And you can explain why a query is slow, what an isolation level really prevents and what happens when the leader dies.