March 13, 2013 | Written by: Piotr Pietrzak
Share this post:
Virtualization of disk subsystems follows the trend of server virtualization.
The technology users notice that virtualization of just the server environment is not enough. If you want to get a high-performance, scalable and highly automated IT solution, virtualization of all the layers is a mandatory requirement.
This virtualization includes servers, network, disk and even tape subsystems. Only full virtualization enables truly flexible allocation of resources, easy reconfiguration, and simplification of the high availability (HA) and disaster recovery (DR) environment for operating under the supervision of the application.
An additional advantage is the possibility of a very thorough investigation of the use of company information resources, and then assigning the resources to the appropriate cost center. It is also a necessary task on the road to building solutions under the “cloud computing” mark, for which full virtualization of resources is a fundamental and indispensable step.
Better use of shared resources brings with it potential cost cuts in both capital and operational expenditures. A shortened response time to reported IT-services demand, concerning the processing power of many repetitive tasks, can be fully automated.
Virtualization of storage subsystems is a natural consequence of server virtualization and the answer to the growing demand for storage. In the case of centralized processing and data storage, the demand for storage space is growing rapidly.
Data is migrated from the distributed insular based systems into centralized resources, data centers, or both. Virtualization increases the effective use of existing storage space and improves control over its allocation. In addition, the physical separation into layers becomes completely transparent for the user and independent from application or concrete devices. The next step is storage tiering, which means division of devices into data storage layers.
With the use of the virtualization technology, tiering becomes transparent for the user and the application. The individual layers can use separate media depending on the nature of the information stored, the required target performance, and price. This approach allows tight control over both the lifetime and the cost of information storage. Data over time can be migrated into slower and cheaper carriers and stays visible and accessible for the user.