🤖 AI Summary
Existing non-transactional consistency levels in distributed cloud object storage are defined heterogeneously, hindering systematic comparison and formal verification.
Method: This paper proposes a unified framework based on the Shared Object Pool (SOP) model, centering on “order” and introducing a novel dual-constraint formalization that jointly captures convergence (via lineage shape) and relational ordering (via operation-relative positions).
Contribution/Results: The framework concisely characterizes essential distinctions among mainstream consistency models—including eventual, causal, and read-committed consistency—and establishes, for the first time, their theoretical availability upper bounds. It further provides rigorous semantic mappings of real-world cloud storage protocols, including Amazon S3 and Azure Blob Storage. The resulting framework offers a formally verifiable, engineering-friendly theoretical foundation for the design, verification, and optimization of cloud object storage systems.
📝 Abstract
We present a summary of practical non-transactional consistency levels in the context of distributed data replication. Unlike prior work, we build upon a simple Shared Object Pool (SOP) model and define consistency levels in a unified framework centered around the concept of ordering. This model naturally reflects modern cloud object storage services and is thus easy to understand. We show that a consistency level can be intuitively defined by specifying two types of constraints on the validity of orderings allowed by the level: convergence, which bounds the lineage shape of the ordering, and relationship, which bounds the relative positions between operations. We give examples of representative protocols and systems, and discuss their availability upper bound.