Why can't we just use wall-clock timestamps to order events across machines?
Clock skew is inevitable in distributed systems. Two machines' clocks drift by milliseconds or seconds, so an event that happens LATER can receive an EARLIER timestamp. Last-write-wins databases relying on timestamps will silently discard the newer write and lose data.
Answered in
Vector Clocks: Detecting Causality in Distributed SystemsVector clocks solve distributed ordering: when wall-clock timestamps fail to detect concurrent writes, vector clocks reveal true causality and conflicts.
Read the full analysisOther questions this article answers
More system design questions
- Why does an index's internal data structure matter if it all ends up 'faster than a scan'?
- Why is disk I/O the thing index structures are actually optimizing for?
- Why can't a hash index handle range queries?
- What makes a bitmap index different from a B-tree, and when is it better?
- Why do B-trees stay balanced automatically as data is inserted?
- Why does a database need an index at all — why can't it just scan the table?
- When is a hash index better than a B-tree index?
- What is a composite index and why does column order matter?
Every answer on Crashtech is written by the editor of the article it comes from — never auto-summarised. Browse all answers or the System Design beat.