Use of Time in Distributed Databases (part 2): Use of logical clocks in databases

submited by

Style Pass

2024-12-26 23:00:07

This is part 2 of our "Use of Time in Distributed Databases" series. We talk about the use of logical clocks in databases in this post. We consider three different approaches:

In the upcoming posts we will allow in physical clocks for timestamping, so there is no (almost no) physical clocks involved in the systems in part 2.

Dynamo employs sloppy quorums and hinted hand-off and uses version vector (a special case of vector clocks) to track causal dependencies within the replication group of each key. A version vector contains one entry for each replica (thus the size of clocks grows linearly with the number of replicas). The purpose of this metadata is to detect conflicting updates and to be used in the conflict reconciliation function. Dynamo provides eventual consistency thanks to this reconciliation function and conflict detection by version vectors.

Cassandra, which provided an opensource implementation of Dynamo, decided to forgo vectors clocks in favor of using physical time supplied by the client and Last-Writer-Wins rule for updating replicas.