Reading list: distributed systems

Notes to myself, mostly so I stop re-googling the same three explanations. These are the ones that actually changed how I build things.

Consensus

• Raft — the paper is readable in an evening and worth rereading.
• Paxos Made Simple — the title is optimistic, but it helps.

Systems in practice

• Designing Data-Intensive Applications — the one book I recommend without caveats.
• The Dynamo paper — for understanding eventual consistency tradeoffs.

Time and order

“Time, Clocks, and the Ordering of Events” is old and still the clearest framing of why distributed ordering is hard.

Most production incidents are a disagreement about what “now” means.

I keep a running list and prune it every few months. Better to reread five things I understand than skim fifty I do not.