CRDTs: From sequential to concurrent executions
When nodes are spread across large geographic distances, being available for local users, and providing short response times, is often at odds with keeping strong consistency across the whole system. Several systems, that target large scale geo-replication, support multi-master operation and transient data divergence, allowing each site to update replicas with no immediate coordination. From the user application perspective, the system cannot be seen anymore as a single sequential copy, since now operations can be processed concurrently at different locations. Conflict-free Replicated Data Types (CRDTs) can take away a lot of the complexity when migrating from a sequential to a concurrent setting.
In this talk we will explore a bit of the path in this transition, cover what can be expected, and present a few guiding principles. Concurrent behaviour will be explored for common data types, such as counters, registers, sets and sequences.
OBJECTIVES
This talk should help the audience understand how to reason about the behaviour of replicated data types that can evolve concurrently, and how that behaviour relates to the familiar sequential execution that occurs in a single centralized copy.