ArangoDB Further Powers Graph at Scale with Release of ArangoDB 3.9

Latest release improves performance across distributed data; also includes new search and query functionality to support analytics use cases and international adoption

San Francisco and Cologne, Germany – February 15, 2022 – ArangoDB, the most scalable open source graph database, today announced the GA release of ArangoDB 3.9. ArangoDB 3.9 includes new capabilities that improve graph query performance across distributed data sets, as well as additional search functionality, supporting the ever-growing need for businesses to perform analytics focused around graph and full-text search.

Key new features in ArangoDB 3.9 include Hybrid SmartGraphs, making it even more efficient to query sharded datasets, as well as additional Analyzers for ArangoSearch, ArangoDB’s natively-integrated, full-text search and ranking engine. IDC’s Research Vice President of Data Management Software Carl Olofson regards the graph database as the next truly revolutionary database management technology, and expects the market to grow 600% over the next 10 years[1]. The release of ArangoDB 3.9 continues ArangoDB’s goal to serve as the most scalable graph database by making it easy to work with any data of any kind.

“Graph technology continues to solve some of today’s toughest business problems, such as supply chain management, fraud detection, and route optimization, to name a few,” said Jörg Schad, PhD, CTO at ArangoDB. “Historically, the adoption of graph databases has been hindered due to difficulties around operating them at scale. We’ve made great strides with the past few releases of ArangoDB to introduce new features that improve performance across large data sets, and with the introduction of Hybrid SmartGraphs, ArangoDB 3.9 is yet another step in this direction.”

Hybrid SmartGraphs (ArangoDB Enterprise)

ArangoDB 3.9 introduces Hybrid SmartGraphs, an extension of ArangoDB Enterprise Edition features SmartGraphs and SatelliteCollections. SmartGraphs allow ArangoDB to scale ‘smartly’ by minimizing network hops across distributed data. Hybrid SmartGraphs take this a step further by combining this functionality with SatelliteCollections, which enable faster join operations with sharded datasets. By allowing queries to be executed locally as much as possible, Hybrid SmartGraphs makes it even easier to scale data effectively and execute queries as efficiently as possible.

New ArangoSearch Analyzers: Segmentation and Collation
ArangoDB 3.9 includes the addition of two new Analyzers that bolster ArangoSearch’s already extensive support for a wide range of languages. The segmentation Analyzer facilitates mixed-language strings, and the collation Analyzer ensures the rules of the respective language are followed. These additional Analyzers assist ArangoDB’s increasing global growth.

AQL: Decay and vector functions, PRUNE variables
While ArangoDB 3.9 includes many new AQL features and improvements, there are three main ones: decay and vector functions, as well as PRUNE variables.

The three new decay functions, exp, linear, and gauss, calculate a score with a function that decays depending on the distance of a numeric value from a user-given origin. These are especially helpful in analytics scenarios or when dealing with time-series data that includes measurements which change over time.

The three new vector functions, cosine similarity, Manhattan distance, and Euclidean distance, can calculate how similar vectors are. This is useful for grouping similar documents together, powering use cases such as text analytics and recommendations.

ArangoDB 3.9 also builds on the PRUNE functionality with the introduction of PRUNE variables. PRUNE reduces the overhead of graph traversal queries by stopping a query when it reaches specific conditions. PRUNE variables adds the option to store a PRUNE expression as a variable, allowing it to later be used in a query without having to repeat the PRUNE condition.

Extended database naming conventions
In order to better support ArangoDB’s international user base, ArangoDB 3.9 accepts database names that include most UTF-8 characters, such as Japanese or Arabic letters, emojis, and letters with accentuation. In addition, many ASCII characters that were formerly banned in the traditional naming convention are also now accepted.

Additional features in ArangoDB 3.9 include various performance and UI improvements.

ArangoDB 3.9 is available immediately for download here, as well as on ArangoDB ArangoGraph, ArangoDB’s managed service.

Learn more


About Arango 

Arango gives you a trusted data foundation for Contextual AI — transforming enterprise data into a System of Context that truly represents the business, so LLMs can deliver better outcomes with unlimited scale and cost efficiency.

The Arango AI Data Platform gives developers a single, integrated environment to build and scale AI-powered applications without the complexity of stitching together multiple databases and tools. At its core is a massively scalable multi-model database that unifies graph, vector, document, and key-value data with full-text, geospatial, and vector search, creating the System of Context—the bridge between enterprise data and AI-powered applications.

The Arango AI Suite includes automated data pipelines, multimodal data ingestion, AIOps and MLOps, LLM integrations, Graph Analytics, agentic frameworks for context-aware Hybrid/GraphRAG, GraphML, natural-language support, and GPU acceleration, enabling repeatable ROI and faster innovation.

Trusted by NVIDIA, HPE, the London Stock Exchange, the U.S. Air Force, NIH, and Articul8, Arango powers enterprise AI with context, confidence, and scale.

Learn more at arango.ai, LinkedIn, YouTube, and G2.
marketing@arangodb.com