/
lectern.tv
Play15s30s1x0:00/--:--
DoneWatched
Expand
lectern.tv
15:30/18:44
3:07
0:35
1:15
2:50
1:08
1:03
2:13
2:19
1:57
2:15
Linearizability
Defines linearizability as a strong consistency model for distributed systems, illustrating its strict real-time ordering guarantee and contrasting it with weaker consistency properties.
Lecture guide · Written by AI from the lecture. For the lecture’s own words, open the Transcript tab.
Defines linearizability as a strong consistency model in distributed systems and compares it with weaker consistency properties. It uses execution timelines to explain how quorum reads and writes fail to guarantee linearizability unless augmented with read repair, and demonstrates how total order broadcast can be used to implement linearizable compare-and-swap operations.
This lecture assumes
- Distributed Systemstaught in Introduction & MapReduce at 0:18
- Transactionstaught in SQL & Relational Databases at 2:05:56
Key Takeaways
- Linearizability provides the illusion of a single, non-replicated, atomic copy of data to clients.
- Real-time ordering dependencies in linearizability are distinct from message-passing happens-before relationships.
- Standard quorum reads and writes are insufficient for linearizability without additional read repair steps.
- Compare-and-swap operations can be made linearizable in distributed systems by building upon total order broadcast.
Concepts Explained5
Linearizability—A consistency model where the system behaves as if there was only a single copy of the data and all operations happen atomically on that single copy.
Serializability—A form of isolation between transactions ensuring transactions have the same effect as if they were executed in some serial order.
Read repair—A mechanism where a client detecting an outdated replica during a read operation updates it by sending the latest value to replicas that missed it.
Blind write—A write operation that unconditionally overwrites the current value of an object regardless of its previous value.
Examples, Proofs & Derivations