What’s fast, is chockablock with goodies and shows up right before the holidays?
SingleStore’s winter release, of course.
SingleStore 8.9 has officially arrived, marking a significant milestone in our journey to redefine real-time analytics and operational performance. Packed with a wide array of innovative features and performance enhancements, this release showcases our commitment to meeting diverse data workload demands based on our customers’ feedback. A heartfelt thank you to our incredible team of engineers for delivering this release as part of the holiday cheer.
How we organize our product features
Before diving into the features, let’s explore how we organize innovations into three categories to clarify our focus and vision for the future:
- Core. Enhancing speed and scalability to empower real-time analytics and operational excellence.
- Connectors. Making it easier for customers to integrate SingleStore with their ecosystem, enabling efficient real-time and AI application development.
- Platform. Advancing features within our cloud-managed service, SingleStore Helios®, to bring workloads closer to the data.
Core enhancements
In 8.9, we’ve focused on both speed at scale and hybrid search capabilities, both of which are critical for building modern AI-first applications. Here are some highlights:
Speed at scale features
- JSON computed columns now load faster, significantly cutting down processing times for high-volume ingestion workloads.
- JSON array operations have been further optimized, with performance improvements up to 20x for aggregation-heavy queries — especially helpful for analytics on nested datasets.
- Disk spilling now supports right and full joins, improving the reliability of memory-intensive queries.
- Writable views allow developers to update, insert and delete data directly through views, simplifying workflows.
- Vector indexes now work on nullable columns, broadening the applicability of vector search — particularly for AI and machine learning workloads.
Full-text search: Beyond English
- Custom analyzers now support multiple languages, including French, Japanese and Arabic, for more accurate and efficient multilingual search.
- N-grams have been introduced to improve fuzzy and partial matching, making features like auto-complete and typo-tolerant search more effective.
Connector enhancements
Data does not live in vacuums. In fact, quite the opposite — and we recognize our commitment to be more open to data ingestion and egress within SingleStore so companies have optionality and control over their own data. In 8.9 we added enhancements in this area under three items.
Continuous ingest from Iceberg tables
- Earlier this year, we announced integration with Iceberg as the world of data lakes and data warehouses moved toward a common standard to manage their “big data” with multiple tools. This release includes upsert and append modes to simplify keeping data updated without downtime or complex ETL processes.
- In addition, with 8.9, SingleStore now supports compatibility with the Polaris and Hive Iceberg catalogs. This enables the above functionality for these catalogs as well as the GLUE, Snowflake, REST and JDBC catalogs.
Simplified pipelines
- Kafka pipelines now support additional compression formats, including lz4 and snappy, optimizing data transfer.
- S3 pipelines feature improved error handling and overwrite capabilities, reducing interruptions.
Loading data on SingleStore Helios
- A no-code interface simplifies data ingestion from S3, CSV and MongoDB sources, reducing time-to-value for developers. Additional source integrations are planned in upcoming releases.
Platform enhancements
For users of SingleStore Helios, new platform features bring added power and usability for data-driven applications.
Interactive API endpoints in Helios
- Build dynamic API endpoints directly using Jupyter notebooks, bridging analytics with real-time applications. This feature is in private preview, and is already generating excitement among early adopters.
Live dashboards with Jupyter Notebooks
- Create live dashboards within Helios using Jupyter, enabling teams to visualize and act on data insights more effectively. These dashboards are ideal for monitoring real-time application metrics, tracking data ingestion workflows or providing stakeholders with actionable analytics tailored to their needs. This feature is currently in private preview.
Try SingleStore 8.9
SingleStore 8.9 is packed with powerful new features and improvements that further our mission to be the go-to database for real-time analytics and operational workloads. From multilingual search capabilities to simplified data pipelines and enhanced platform features, this release underscores our focus on innovation. Explore the release notes to dive deeper into the technical details, and if you are not yet part of the SingleStore community give it a try with our free-for-life tier.