Appistry is expanding its Cloud IQ platform, a basis for building a private cloud inside the enterprise, to include a file storage system. Cloud IQ Storage is aimed at extremely large storage loads-- up to a petabyte -- that would enable next generation, high scalable applications dealing with masses of data.Appistry announced beta availability of Cloud IQ Storage on Monday, the opening day of Cloud Connect in Santa Clara, Calif. The product will become available sometime in the second quarter, said Kevin Harr, CEO of Appistry, in the announcement.

Cloud IQ for Storage can be used as a stand -alone product or with an earlier Appistry product, Cloud IQ Engine. The two together can be used to store data across a cluster of commodity servers, then call it up and process it on the server where it is resident for distributed problem solving. Such a distributed approach can deal with masses of data by expanding the server cluster without requiring a re-engineering of the application, said Harr.

Instead of moving data from many storage sources into the application, Cloud IQ Storage moves the application close to the distributed data and coordinates the processing. Administrators may specify policies governing the replication of data, allowing a means of achieving distributed fault tolerance.

Cloud IQ for Storage balances disk utilization across a server cluster, across multiple clusters in a data center or across different data centers, Harr said. It is geographically aware and can mirror files in different locations to give the system disaster recovery capabilities.

A second version of Cloud IQ for Storage, Hadoop edition, provides plug and play capability with the open source distributed data management system.

Harr claimed the Hadoop edition avoids any single point of failure or network bottleneck when working with the Hadoop Distributed File system. It can be deployed in place of the Hadoop file system where reliability and throughput are key attributes, he said. The Hadoop edition ships with Hadoop Distributed File System drivers.

Yahoo is a user of the Hadoop open source and contributor to the project. Hadoop employs MapReduce, a way of linking data to its closest processing element in large server cluster, and its distributed file system. MapReduce was created inside Google for dealing with very large data sets distributed across commodity servers.

Appistry is positioning its Cloud IQ platform as an enhancer and extender of Hadoop-like capabilities. No pricing has been announced.

InformationWeek has published an in-depth report on cloud computing and service-level agreements. Download the report here (registration required).