Original URL: http://www.theregister.co.uk/2012/07/25/dell_rainstor/
Dell channels Dr Who, shoves Big Data into Small Data Tardis
RainStor to play the time lord's photogenic assistant
Dell reckons it can turn Big Data into, er, Small Data. Hoarded information could be squeezed down to size - potentially reducing disk occupancy by 97 per cent - if the tech titan's numbers add up.
The company will bundle RainStor's deduping structured data software with one of its object storage system to provide a deduplication ratio of up to 40:1. Before you declare the Reg storage desk to be staffed by fools, it's Dell is saying this, not us.
The Texan firm chose to use its DX6000 Caringo-based object store, which can scale up to petabytes of capacity and billions of objects.
The combination is called Dell's Big Data Retention product; actually it's termed a "solution" but damn it, it's a product. The reduction magic is in the RainStor technology. This deduplicates database information and gets highish dedup rates from doing so. This is especially the case with records containing a high degree of repetition such as, for example, telephone company call logs.
RainStor differs from Dell's Ocarina data reduction technology in that Ocarina is not suited to structured information and RainStor isn't focussed on unstructured and semi-structured data, such as images.
Recently RainStor added the capability for its analytics engine software to work directly on Hadoop files and run either MapReduce queries or its own. This must have resonated with Dell, which says its Big Data Retention products can store structured, semi-structured and unstructured data from an unlimited number of data sources, formats and types. Its single and massively deduping repository is said to provide cost-savings in storage and management.
Dell says its retention bundle can be a front-end repository for large datasets, a back-end archive, a standalone repository, or an analytics platform with Hadoop.
The RainStor software can be installed on any combination of Dell's servers and storage, having, Dell says, unlimited scalability and zero administration. It can be used on-premise or in a cloud system. Dell is now able to say it offers a single logical Big Data silo, that, like Dr Who's Tardis, is a lot bigger inside than it appears from the outside because of RainStor data reduction. ®