In this lecture we analyze key-values databases. At first we introduce key-value characteristics, advantages and disadvantages.
Then we analyze the major Key-Value data stores and finally we discuss about Dynamo DB.
In particular we consider how Dynamo DB: How is implemented
1. Motivation Background
2. Partitioning: Consistent Hashing
3. High Availability for writes: Vector Clocks
4. Handling temporary failures: Sloppy Quorum
5. Recovering from failures: Merkle Trees
6. Membership and failure detection: Gossip Protocol