Kognitio has introduced a solution it says will help companies reduce the amount of time required to run ad-hoc queries with Big Data stored in Hadoop, by making Hadoop integration part of the Kognitio Analytical Platform.
According to Kognitio, the solution goes beyond simple connectivity, with an external tables feature that will embed a software agent inside Hadoop clusters. Thus, it effectively leverages the Kognitio platform as an in-memory analytical accelerator for Hadoop that uses industry-standard servers and standard SQL.
The platform’s native, in‐memory analytical processing capabilities can return query results in seconds, which will help users save query processing time. Kognitio says it has achieved load rates of data into memory in excess of 15 terabytes per hour.
The open source Apache Hadoop software library, a framework that enables distributed processing of large data sets across clusters of computers using a simple programming model, is becoming popular among organizations interested in analyzing Big Data. Yet Hadoop requires technical skills not likely to be found on business intelligence teams accustomed to working with the SQL and Multi-Dimensional eXpressions (MDX) languages.
In a company statement, Kognitio CTO Roger Gaskell said, “The Kognitio Analytical Platform is the must-have layer to accelerate Hadoop and allow it to work with standard BI and OLAP tools, enabling results that make a difference on the bottom line.”
Kognitio is not the only vendor trying to make Hadoop more accessible to "average" business users. Yesterday Teradata introduced Aster SQL-H software, which it says empowers business analysts to directly analyze Big Data stored in Hadoop without requiring MapReduce development skills or an understanding of how or where data is stored within HDFS.
Kognitio software runs on industry-standard x86 servers, as an appliance or in the Kognitio Cloud, a private or public cloud platform-as-a-service (PaaS) that is used to deliver the Kognitio Analytical Platform on a subscription basis.