Distributed SQL: Features, Core Characteristics, and Cloud-Native Requirements

As organizations migrate their workloads to cloud environments, traditional relational databases often limit migration speed and flexible scaling, prompting a desire to retain the reliability of systems like Oracle, SQL Server, Postgres, and MySQL while leveraging cloud scale and global stability.

To meet these needs, many turn to NoSQL databases, but because NoSQL is not designed for full ACID consistency, it cannot serve transactional workloads requiring strong isolation for tasks such as financial accounting, inventory control, and identity management.

Distributed SQL – A New Kind of Database

In 2012, Google published a paper on Spanner, introducing a globally scalable, distributed, multi‑version, synchronously‑replicated database that was the first to support externally‑consistent distributed transactions worldwide.

Building on this foundation, the article discusses the basic concepts of distributed SQL, focusing on how to achieve scalability and consistency. Distributed SQL databases typically exhibit seven core characteristics:

1. Scalability

Like elastic compute, distributed SQL databases can scale seamlessly in cloud environments without added operational complexity, distributing data evenly across multiple participants.

2. Consistency

They must provide high isolation in distributed settings, handling resource contention and delivering transaction isolation comparable to single‑instance databases.

3. Elasticity

Without external tools, they offer top‑level resilience, automatically replicating data and minimizing recovery time in continuously online cloud services.

4. Geo‑Replication

Leveraging cloud services, they break geographic limits, enabling distributed processing and storage to meet user demands worldwide.

5. SQL Support

SQL remains the universal language for applications; distributed SQL solutions such as Spanner, Amazon Aurora, Yugabyte, FaunaDB, and CockroachDB all support it, avoiding the need for retraining developers.

6. Data Locality

By partitioning data based on fields, distributed SQL can place data closer to users, addressing data sovereignty concerns, reducing latency, and lowering cloud transfer costs.

7. Multi‑Cloud Mode

Each semi‑autonomous unit can be deployed independently and join larger clusters (e.g., CockroachDB), allowing deployment across any public, private, or on‑premise cloud without restrictions.

Basic Requirements of Distributed SQL

Beyond the seven features, a distributed SQL database must also provide fundamental database capabilities:

Manageability: command‑line or GUI tools for installation, configuration, lifecycle management, backup, restore, schema definition, indexing, partitioning, and DDL operations.

Optimizable: advanced features like cost‑based optimizers for query performance tuning.

Security: authentication, authorization, accountability (AAA), and integration with centralized identity and governance systems for consistent data policies.

Integrability: support for tested drivers and ETL tools to connect with front‑end applications and downstream services.

These basic functions aim to meet the demands of enterprise‑grade applications.

Conclusion

Distributed SQL databases represent an emerging class that must continue improving consistency and data locality within cloud environments to address performance and efficiency challenges in production workloads.

CockroachDB, a cloud‑native distributed SQL database, exemplifies these concepts and can serve as a practical entry point for enterprises looking to migrate workloads to the cloud.