Announcing ScyllaDB 2025.4, with Extended Tablets Support, DynamoDB Alternator Updates & Trie-Based Indexes

An overview of recent ScyllaDB changes, including extended tablets support, native vector search, Alternator enhancements, a new SSTable index format, and new instance support The ScyllaDB team is pleased to announce the release of ScyllaDB 2025.4, a production-ready ScyllaDB Short Term Support (STS) Minor Feature Release. More information on ScyllaDB’s Long Term Support (LTS) policy is available here. Highlights of the 2025.4 release include: Tablets now support Materialized Views (MV), Secondary Indexes (SI), Change Data Capture (CDC), and Lightweight Transactions (LWT). This fully bridges the previous feature gap between Tablets and vNodes ScyllaDB Vector Search is now available (in GA), introducing native low-latency Approximate Nearest Neighbor (ANN) similarity search through CQL. See the Getting Started Guide and try it out. Alternator (ScyllaDB’s DynamoDB-compatible API) fully supports Tablets by following tablets_mode_for_new_keyspaces configuration flag, except for the still-experimental Streams. The new Trie-based index format improves indexing efficiency. New deployment options with i8g and i8ge show significant performance advantages over i4i, i3en as well as i7i and i7ie For full details on how to use these features — as well as additional changes — see the release notes. Read Release Notes Vector Search Vector Search Support ScyllaDB 2025.4 introduces native Vector Search to power AI-driven applications. By integrating vector indexing directly into the ScyllaDB ecosystem, teams can now perform similarity searches without moving data to a separate vector database. CQL Integration: Store and query embeddings using standard CQL syntax. ANN Queries: Support for Approximate Nearest Neighbor (ANN) search for RAG and personalization. Dedicated Service: Managed vector indexing service ensures high performance without impacting core database operations. Availability: Initially launched on ScyllaDB Cloud. For more information: ScyllaDB Vector Search: 1B Vectors with 2ms P99s and 250K QPS Throughput Building a Low-Latency Vector Search Engine for ScyllaDB Quick Start Guide to Vector Search Extended Tablets Support The new release extends ScyllaDB’s tablet-based elasticity to use cases that involve advanced ScyllaDB capabilities such as Change Data Capture, Materialized Views, and Secondary Indexes. It also extends tablets to ScyllaDB’s DynamoDB-compatible API (Alternator). Alternator Improvements Alternator, ScyllaDB’s DynamoDB-compatible API,  now more closely matches DynamoDB’s GetRecords behavior. Event metadata is fully populated, including EventSource=aws:dynamodb, awsRegion set to the receiving node’s datacenter, an updated eventVersion, and the sizeBytes subfield in DynamoDB. Performance was improved by caching parsed expressions in requests. That caching reduces overhead for complex expressions and provides ~7–15% higher single-node throughput in tested workloads. Alternator also adds support for per-table metrics (for additional insight into Alternator usage). Trie-Based SSTable Index Format A new trie-based SSTable index format is designed to improve lookup performance and reduce memory overhead. The default SSTable format remains “me,” but a new “ms” format is available, which uses trie-based indexes. The new format is disabled by default and can be enabled by setting sstable_format: ms in scylla.yaml. When enabled, only newly created SSTables use the trie-based index; existing SSTables keep their current format until rewritten with nodetool upgradesstables. New Deployment Options This release expands support to all I7i and I7ie instance types (beyond the previously supported i7i.large, i7i.xlarge, i7i.2xlarge). These instances offer improved price-to-performance compared to previous-generation instances. Support was also added for the i8g and i8ge families, which provide better price-to-performance than x86-based instances. Read the Release Notes for More Details