Graph databases are finding growing use in enterprise analytics projects thanks to their ability to efficiently keep track of the relationships among records. And no system in the category is more popular these days than Neo4j, which received a major update this morning that aims to expand its appeal even further.

The arguably most important enhancement in the release affects the way that servers are synchronized within large-scale deployments. Neo4j 3.1 replaces the nearly 20-year-old Paxos algorithm that was used to handle the task in previous versions with a much newer alternative called RAFT. A staffer for Neo Technology Inc., the company behind the graph store, explained in a blog post last year that the protocol is easier to implement and thus reduces the room for errors. The other big benefit is that it supports a wider range of cluster configurations, which should give users more operational wiggle room.

And in the same spirit, Neo also expanded its database’s hardware support by adding compatibility with IBM Corp.’s CAPI Flash interconnect. The latter vendor claims that the technology can move bits between its POWER8 chips and memory up to twice as quickly as rivaling offerings from Intel Corp., which allows database operations to be executed noticeably faster. As a result, the functionality should make Neo4j much more appealing to the growing numbers of companies that using servers based on Big Blue’s processor series.

Rounding out the new release is a set of security features designed to help organizations regulate the information in their deployments more effectively. Neo4j 3.1 can plug into a company’s user directory to provide authentication features, packs a natively encrypted command line integration and introduces numerous other smaller enhancements as well. Neo hopes that the addition will make its system better-suited for use in industries such as healthcare and finance that have strict data privacy requirements.

The new functionality complements the capabilities that the company introduced in the 3.0 release of the database earlier this year. The most important addition was a homegrown binary protocol called Bolt that makes it possible to pull information from Neo4j  much faster than HTTP, thus speeding up queries. Neo can be expected to come up with many more such value-added features over time as part of its efforts to maintain its lead in the graph database segment.

Image via Pixabay