ScalarDB 3.19 has been released

Database

Microservices

Product

Scalardb

ScalarDB 3.19 has been released!

Josh Wong

4 min read·

3 days ago

--

We have recently released ScalarDB 3.19, a new minor version of ScalarDB. This release significantly strengthens data consistency management in microservices architectures, particularly in the age of AI-driven development.

Version 3.19 includes new features and improvements, including:

In this post, we’ll introduce these new features and improvements.

ScalarDB Saga for handling complex business workflows

With the rise of microservices and AI-driven development, enterprises are building systems with distributed services that often interact with external APIs, SaaS platforms, and asynchronous processes. While 2PC provides strong consistency for multi-database transactions, it isn’t well-suited for workflows that span services beyond your control or involve long-running business processes.

ScalarDB 3.19 introduces ScalarDB Saga, a new component that orchestrates the Saga and Try-Confirm/Cancel (TCC) patterns for exactly these scenarios. ScalarDB Saga manages distributed workflows that span multiple services, including those that involve:

The Saga component maintains centralized control over workflow execution and state transitions, handling retries and compensation when failures occur — all without requiring complex multi-phase coordination logic in your application code. This means your teams can focus on business logic while the platform ensures data consistency across your entire microservices architecture.

Enterprises can now choose the right consistency pattern for each use case: 2PC for strong consistency across internal databases, and Saga/TCC for flexible orchestration across external systems and asynchronous workflows.

For more information, see the README in the ScalarDB Saga repository.

Microservices architectures often span multiple ScalarDB Cluster instances, each potentially serving different business domains or availability zones. When executing 2PC transactions across these clusters, developers previously had to implement complex multi-phase coordination in application code — handling commit phases and implementing sophisticated error handling for network and service failures.

ScalarDB 3.19 introduces a new transaction interface that abstracts away this complexity. Applications can now execute distributed transactions across multiple ScalarDB Cluster instances by using a simple, single-phase interface. ScalarDB internally manages the full 2PC protocol, coordination, and failure recovery — all transparently.

By removing the mechanical complexity of 2PC coordination from application code, developers can focus on business logic while maintaining strong consistency guarantees across your data infrastructure.

For more details, see the Microservice Transaction API section of the ScalarDB Cluster Java API Guide.

When running analytical queries across multiple heterogeneous databases, data transfer between ScalarDB Analytics and backend databases can become a bottleneck. Complex queries that filter data might retrieve far more rows than necessary, consuming network bandwidth, computing resources, and increasing query latency.

ScalarDB Analytics previously supported predicate pushdown for JDBC datasources. In ScalarDB Analytics 3.19, this support is extended to the ScalarDB datasource, pushing SQL WHERE clause conditions down to backend databases whenever possible. This means:

By bringing computation closer to the data, ScalarDB Analytics queries execute more efficiently across your entire data landscape — whether your databases span on-premises systems, multiple clouds, or a combination thereof.

For more details on ScalarDB Analytics, see ScalarDB Analytics Design and Implementation.

ScalarDB 3.19 strengthens the data consistency foundation for microservices architectures in the age of AI-driven development. By introducing ScalarDB Saga for complex business workflows, simplifying 2PC coordination, and accelerating analytics queries, ScalarDB 3.19 enables enterprises to:

Whether you’re modernizing enterprise systems, building new microservices, or leveraging AI-driven development practices, ScalarDB 3.19 provides the data consistency and operational capabilities your platform needs.

For details about all changes in this release, please refer to the ScalarDB 3.19 release notes.

To learn more about ScalarDB, please take a look at our other posts on the Scalar Engineering Blog and at the ScalarDB documentation.

If you are considering using ScalarDB, please contact us.