Object Storage

Hybrid Cloud Software-Defined Storage Solutions in HPC, Cloud and Big Data

When applied to big data and technical computing applications, traditional storage systems have inherent weaknesses. These systems cannot scale to massive volumes of data. Petabyte-scale SAN environments are expensive, complex, and hard to manage. While some NAS systems can expand to a few petabytes of capacity, they are unable to cope with millions to billions of file. NAS file systems are engineered to organize content into human understandable hierarchy, with the emphasis on categorization and comprehension not performance.

Object Storage Makes Big Data a Reality

Object Storage systems were designed to overcome these weaknesses, making 21st Century storage for big data a reality. Object storage brings new methods to storing, organizing and accessing data. Rather than using files in a directory hierarchy, data is stored in an object, which also contains metadata describing the type of data, and other pertinent information. Objects are stored in a flat organization of containers. Objects are referred to by their unique ID.

Object storage systems are designed for machine access rather than human access. Thus, the performance is not impacted by the number of objects in the system, nor the actual location of the object. Object storage systems typically protect data by maintaining multiple copies. Should a single drive fail, the data is always available on other drives, and making new copies of the data has little impact on the overall storage system performance.

RAID, Inc. is proud to offer the Ability high density object storage enclosure, enabling petabyte-scale object storage environments while occupying the least amount of expensive datacenter floor space. Using 96 10TB drives, Ability can store (960 TB) in just 4U of rack space.