Press release: Objectivity, Inc. Launches InfiniteGraph, the Distributed Graph Database for the Enterprise

The beta program for InfiniteGraph DB, a NoSQL graph database, capable of distributed and virtually unlimited scalability, enables developers to build and deploy highly complex social networking and intelligence applications.

Objectivity, Inc., a leading provider of data management solutions, today announced that its leading-edge distributed graph database product for the enterprise, InfiniteGraph, is now available in beta. Developed by Objectivity’s new InfiniteGraph business unit, the InfiniteGraph distributed database technology and developer application programming interface (API) were created to help organizations obtain real-time answers from deep analysis of complex, multi-dimensional relationships that live within their mission-critical data.

InfiniteGraph’s database and API has been designed to make it easy for developers to exceed the highest level of graph computing requirements and to future-proof applications that may need to scale significantly in search of billions of connections and relationships in massive volumes of complex data. Traditional database systems often require months of extra development, resource intensive customized code, and costly middleware layers to try and store and find connections in their disparate data systems. In comparison, InfiniteGraph represents the next-generation of graph database technology that provides developers with the tools to support the ultimate combination of simplicity and scalability, enabling organizations to find, store and exploit the relationships hidden in their data at any time.

“The InfiniteGraph solution will significantly reduce the technical barriers to advanced relationship analysis, allowing organizations to utilize their complex and distributed data in near real-time, and on virtually any distributed or cloud-based computing platform,” said Darren Wood, chief developer for the InfiniteGraph team. “InfiniteGraph also supports the “NoSQL” (or “Not-Only-SQL”) community, which seeks to promote the best solutions for various problem sets, which cannot be satisfied by traditional SQL-based technologies.”

InfiniteGraph’s beta period will last approximately six weeks, concluding with the general release of the product in mid-July. During the beta period, the developer community is encouraged to visit the InfiniteGraph web site and participate in the beta program, adding to the public discussions and providing input that will further the development of the product. The company will also be engaged with the InfiniteGraph community through various social media channels including Twitter, Facebook and LinkedIn.

“InfiniteGraph has been designed to exceed the technical and business requirements for a wide range of audiences and end-users — from individual developers and emerging startups, to the largest enterprise and government entities imaginable.” said Jay Jarrell, president and CEO of Objectivity. “We’re offering a public beta program to give the entire developer community and graph ecosystem the opportunity to contribute to a product that addresses their demands for distributed and highly scalable graph processing solutions.”

The InfiniteGraph distributed graph database will provide a future-proof solution for organizations seeking to navigate through the volumes of data they collect every second to identify complex relationships and connections. The technology is highly flexible and supports a wide range of usage and deployment options, whether as the primary database management system, or as an enhancing extension to existing relational infrastructures, map reduce and batch processing layers. InfiniteGraph was also designed to require virtually zero administration, so whether it is deployed in a corporate data center, globally distributed cloud platforms, or even as an embedded data component within countless sensors and other devices, InfiniteGraph supports the most demanding requirements in applications and systems that simply must work all the time.

The InfiniteGraph API will simplify the way that developers create sophisticated, next-generation solutions. These next-generation platforms and applications will empower organizations involved in areas including social networking, business intelligence, scientific research and national security to get instant answers and information from vast amounts of highly complex and often unstructured data.

“The next generation of data management solutions include graph databases, such as InfiniteGraph, which are becoming very significant in several areas, particularly those related to complex relationship analysis.” says Carl Olofson, research vice president of Database Management and Data Integration software research at IDC. “Objectivity’s early entry in this space with the InfiniteGraph solution is sure to garner a great deal of attention from the NoSQL and graph database communities.”

Following the InfiniteGraph beta program, the company plans to offer several options for licensing and support, including the ability for qualified startups and new businesses to use InfiniteGraph completely free in their deployments.

  • Developers can download and try InfiniteGraph free for two months.
  • Qualified start-ups and new businesses with a GoGrid Cloud Hosting account will be able to use InfiniteGraph for free, per an exclusive offer provided by InfiniteGraph and its cloud platform partner, GoGrid.
  • Developer licenses and support options are also available for as little as $999.00 per year.
  • The company also supports virtually any other type of runtime deployment or OEM licensing request as well, providing custom quotes based on the customers’ unique requirements and requested services and support options.

Developers are encouraged to participate in the InfiniteGraph beta program by visiting

Follow InfiniteGraph on Twitter:

Join our LinkedIn network:


  1. Harry
    Posted June 9, 2010 at 4:53 am | Permalink

    Distributed architecture is (normally) great but given a huge and heavily interconnected graph, how do you:

    a) Partition the graph into shards for distribution?
    b) Avoid passing millions of node ids between shard servers when pursuing long paths for matches?

    • Posted June 10, 2010 at 3:10 pm | Permalink

      InfiniteGraph’s product development roadmap includes strong focus on specific processing per node based on a partition of the graph. We are currently working on:

      1) A best case partitioning with custom placement strategies based on initial ingest. As the graph changes, dynamically reconfigure the vertices to make sure we have a best case partitioning under the new conditions.

      2) A calculated combination of moving data and moving path processing to optimize traversal.

      Graphs are “alive” and often changing, which makes them difficult to perfectly partition. It is sometimes more efficient to simply grab remote data as opposed to transferring the processing of a path to be co-located with it. That is why we have started with an underlying distributed data platform that is built from the ground up to efficiently deal with this exact use case.

  2. Harry
    Posted June 14, 2010 at 7:38 am | Permalink

    Thanks for the reply. It sounds like you are still developing some strategies.
    The fundamental problem I have is that while a distributed approach could be argued to give “infinite” scalability to problems like standard search problems e.g. Google, the suggestion that “infinite” scalability is acheivable for graph queries through distributed systems is misleading.
    The bigger the graph, then typically the more interconnected it becomes. The more interconnected it becomes then the less able to partition into shards.

    In this context the LinkedIn approach of caching the entire graph in RAM seems to make more sense (see )
    Of course this is not “infinitely scalable” either but at least they don’t suggest that this is acheivable.

    • Posted June 25, 2010 at 3:50 pm | Permalink

      One key to the scalability is the single logical view that the InfiniteGraph object identifier provides. Connections exploit those identifiers, so a graph can span multiple logical and physical locations.

      There is no need for servers to share information between themselves, because the kernel within the client figures out where a required edge or vertice is and communicates with the data server closest to the actual data. It’s worth noting that the DBMS that underpins InfiniteGraph has been used to build petabyte+ databases with trillions of highly interconnected objects. It is unlikely that users will get close to its limitations in the near future and, if they do, the address space can easily be expanded.

Post a Comment

Required fields are marked *

%d bloggers like this: