Appistry, a cloud computing company based out of St. Louis, has announced the beta release of a new product line targeting the vibrant web scale storage market. The offering dubbed Cloud IQ Storage, aims to appease the ever growing storage appetite of organizations to deliver web scale storage for a wide range of workloads and use cases.
Web scale storage is becoming quite crowded with many offerings out there, each wit their own unique twist on featureset, delivery model, and target audience. For example of two such players, Mezeo targets service providers with a "cloud storage in a box" solution, while MaxiScale provides massive scale storage with speedy storage retrieval/serving. Appistry is targeting the what they call "computational storage".
Their pitch and vision of computational storage mirrors what is delivered in a MapReduce paradigm where it is more efficient to move your programs and processes to where the data is located instead of moving the data or accessing it over the network. With this vision is it not surprising that Appistry is also delivering their Cloud IQ Storage solution in a secondary flavor dubbed Hadoop Edition that can run in place of the traditional Hadoop Distributed File System (HDFS).
With organizations like Yahoo! and Facebook building their companies intelligence platforms on top of Hadoop, and firms like Hadoop focused startup Cloudera, it makes sense to target that ever growing Open Source Hadoop community. Where Cloud IQ Storage excels over the traditional HDFS setup is in its resiliency.
Where HDFS, like most solutions, uses a central location dubbed a name-node, to store the locations of all of a systems files, IQ Storage distributes the concept of a name-node across all the machines that are storing the files. This alleviates the issue with the name-node being the single point of failure.
This Hadoop maneuver will not only allow Appistry to draft behind Hadoop adoption, but also in a small percentage of customers, use the storage offering as a "gateway-tech" of sorts to get them to adopt their othe offerings such as Cloud IQ Engine which manages computational resources in a similar fashion to the Hadoop MapReduce paradigm.
Those interested can check out Appistry’s site here and contact them for more information.