Consistency-Availability Trade-offs in Distributed NoSQL Systems: From the CAP Theorem to Tunable and Strongly Consistent Cloud Data Stores
Naser Mislimi1*, Violeta Manevska2
Abstract
The move of data management from centralized relational servers to geo-distributed, partitioned platforms forced designers to face a trade-off that classical transaction theory set aside: when the network fails, a replicated system cannot stay both consistent and available. This paper examines how the consistency–availability trade-off in distributed NoSQL systems, first stated in the CAP theorem and later refined by PACELC, has shaped two decades of NoSQL and NewSQL design. Building on the distributed database and NoSQL treatment of Elmasri and Navathe, the paper organizes the consistency spectrum from linearizability through causal and session guarantees to eventual consistency and traces it in eleven representative systems: the Dynamo lineage, Cassandra’s tunable consistency, and strongly consistent designs such as Spanner, Calvin, CockroachDB, and FoundationDB. Two research tables summarize the model hierarchy and a PACELC classification of the surveyed systems. The analysis shows that the field has not converged on one point in the design space; it has industrialized the whole spectrum and exposed consistency as a per-operation, application-controlled parameter. The paper closes with four open problems: accurate time, the industrial absence of causal consistency, semantic verification, and elasticity versus coordination. The decisive question for designers is no longer consistency or availability, but how much latency, complexity, and money a given guarantee is worth.
Keywords
NoSQL databases; distributed databases; CAP theorem; PACELC; eventual consistency; quorum replication; NewSQL; cloud data stores
Cite This Article
Mislimi, N., Manevska, V. (2026). Consistency-Availability Trade-offs in Distributed NoSQL Systems: From the CAP Theorem to Tunable and Strongly Consistent Cloud Data Stores. International Journal of Scientific Advances (IJSCIA), Volume 7| Issue 5: Sep – Oct 2026, Pages 683-691. URL: https://www.ijscia.com/wp-content/uploads/2026/10/Volume7-Issue5-Sep-Oct-No.1089-683-691.pdf
Volume 7 | Issue 5: Sep – Oct 2026
