MemSQL 2.5 "Fuses" Structured and Semi-Structured Data in Single Database

MemSQL announced that its distributed in-memory database can now provide JSON analytics, to deliver a consolidated view of structured and semi-structured data, including standard enterprise and social-media data.

“We provide a scalable, dynamic SQL in-memory database. It is a very fast system,” said Eric Frenkiel, CEO, MemSQL.  “It is fully relational so you can extract any analytical query in the standard SQL that is used in the enterprise. The benefit of using our system is that it offers massive concurrency enabling the ingestion and extraction of data simultaneously. This is effectively a hybridization of OLTP and OLAP in one database which has typically been on separate spheres in the past.”

The overarching theme for the MemSQL 2.5 release is: how do you add unstructured or semi-structured data inside of the system, and that is what the new biggest feature of support for the JSON data type addresses, explained Frenkiel. The new capability allows organizations to combine two disparate data sources for operational analytics, network security, real-time recommendations, and risk management.

While JSON (Java Script Object Notation) has become a popular syntax for storing and exchanging semi-structured data from social-media networks, this information has been largely untapped in conjunction with structured data, leaving companies with an incomplete view into their entire customer bases, says MemSQL.  According to the vendor, existing NoSQL databases offer support for JSON and the querying and parsing of JSON structures, but do little in the way of real-time analytics. Complicating matters, the company contends, is the fact that SQL has no native support for JSON, which makes it difficult to query data across these popular data types. This lack of integration limits an organization’s ability to maximize the power of real-time, big data analytics. With the addition of JSON analytics, MemSQL aims to consolidate the database market and eliminate the need for middleware solutions, enabling companies to realize the benefit of lower total cost of ownership and achieve efficiency gains by eliminating the difficulty of moving data around.

“This basically fuses SQL and semi-structured data into one system,” said Frenkiel. “Building in the JSON support lets our customers now query that JSON arbitrarily and traverse the JSON ad hoc - and they can do so through pure SQL context.”

For more information, please visit