← Library / Distributed Systems & Fundamentals
Slide 1Slide 2Slide 3Slide 4Slide 5Slide 6Slide 7Slide 8Slide 9Slide 10
1 / 10

Use the arrows, swipe, or press ← → to move through the slides.

Distributed Systems & FundamentalsCarousel · 10 slidesSystem Design Series · Day 6

Distributed Transactions — 2PC, Sagas and exactly-once semantics

How multiple services stay consistent when any of them can fail: two-phase commit, the Saga pattern, idempotency and exactly-once delivery.

The trade-off I'd flag

Two-phase commit is simple to reason about and terrible to operate at scale. Sagas push the complexity into compensating actions — which is where it belongs for long-running business flows.

← PreviousData Replication — sync vs async, eventual consistency, conflicts Next →File System Design — storing and querying 100 TB of event logs