I've Got the Power

June 24, 2011
Like a utility grid, grid-based storage technology allows storage to be provided to any application on an as-needed basis. This technology

Like a utility grid, grid-based storage technology allows storage to be provided to any application on an as-needed basis. This technology understands key characteristics about data and its usage and, therefore, is able to manage data based upon the need for its clinical information.

In addition to intelligence, grid storage provides reliability when accessing medical digital data. It adapts to a changing environment, and healthcare facilities reap cost savings from reduced IT overhead and equipment.

There are different types of grid solutions. Simple storage grids provide an access point, usually a file system, where data can be stored. Behind the scenes, the grid is at work, making sure the storage of that data is optimized.

Intelligent grid solutions are "data-aware." They understand clinical meta-data and content to further optimize storage. This type of awareness is controlled by policies, or storage plans, which are developed by each institution.

Grid storage shares attributes with other common consumer utilities, such as electricity. For example, changes in electrical demand are handled transparently to the users and payment is based on amount used. Provisioning of electricity is available and reliable, and there is redundancy in the system that users never see. In addition, it is standardized: Many different devices can all utilize it and connection is simple.

Grid storage also functions like a utility. On-demand storage means as storage demands change, more storage can be added or removed from the system without interruption. Storage can be added incrementally as it is needed.

Storage grids are reliable and allow for redundancy, automatic failover, and load balancing of storage usage, which is accomplished through virtualization.Through virtualization, the storage grid becomes a single brand, and any application that can write files can utilize the grid. Virtualization also provides for integration, whether through a standard file system, or using healthcare standards such as DICOM and HL7

In a healthcare institution, a storage grid could be implemented in two ways. One is as a hosted storage grid, which contains virtualized storage that is maintained over a wide-area network (WAN) from a third-party data center. An alternate is implementation of storage grid middleware on servers and storage devices that reside within the IT infrastructure. These implementations could also be mixed together, such that the hosted storage grid is implemented for business continuity.

Unique attributes make storage grids particularly relevant in healthcare environments.

One, digital data is growing at healthcare facilities. In radiology, multi-slice CT exam size is approaching 100 megabytes on average, while other digital imaging modalities, including digital cardiology and radiology studies, mammography, MRI, ultrasound, and visible light imaging, are also generating substantial amounts of digital data every day. If all of this data is managed in separate application-specific archives, unneeded IT costs arise.

Two, a storage grid can facilitate the rapid access to critical clinical data, such as access to prior patient exams, and back ups with the extreme reliability built into this type of architecture.

Finally, a storage grid inherently provides not just disaster recovery, as mandated by Health and Insurance Portability and Accountability Act (HIPAA) for adequate disaster recovery for patient medical data, but also business continuity. It is a significant improvement over the current state of disaster recovery in healthcare institutions today, which often involves CDs, tapes, and other storage devices scattered across the enterprise or in remote warehouses. In the case of a real disaster, full recovery from this ad-hoc environment is likely to be impossible.

A storage grid, however, can keep data accessible and recover any lost storage in the background, while the clinical systems that require this type of data continue to operate.

Author Information:Ken Rosenfeld Ken Rosenfeld is worldwide general manager of information management solutions, Kodak Health Group, Rochester, N.Y.