Virtualization is one of the hottest technologies today and all major product vendors offer a variety of solutions that are applicable in different areas. One of the key areas in this domain is data center virtualization, whose vital component is storage virtualization.Storage
There are basically two choices for
storage virtualization — block level storage virtualization or file level storage
virtualization. By talking about the former option for storage virtualization, we refer to storage
area network (SAN) virtualization. While considering the latter option for storage virtualization,
we refer to network attached storage (NAS) virtualization. While storage virtualization brings in
many benefits, there are some challenges as well that the storage
virtualization solution has to address. These include scalability, functionality, manageability
An environment which lacks storage virtualization will have islands of storage (or storage systems) from the same or different vendors. Each storage system caters to requirements of various applications individually as well as in terms of IOPS (Input/output operations per second) and capacity. But after the storage environment is virtualized, it is not possible to individually address storage systems. Instead, the whole environment should be analyzed. Moreover, the implemented and virtualized infrastructure has to be capable of handling individual requirements of applications that it caters to. Hence, while conduction evaluation exercises for storage virtualization solutions, care needs to be taken that they are designed keeping in mind future storage requirements.
Since an environment for storage virtualization will have storage from different vendors or the same vendor, each system will have advanced functionality. Hence, the storage virtualization solution must provide the same functionalities as well as continue to leverage the arrays' existing functionality.
Introduction of virtualization in the storage infrastructure divides the end-to-end view in to three distinct views—server to virtualization device; virtualization device to physical storage; and the virtualization appliance itself. Storage virtualization should address manageability of distinct views as well as integrate with existing management tools to enable end to end management of the virtualized environment. Foreffective storage virtualization, the solution should not bring with it unnecessary complexities to the current storage infrastructure, as that will make manageability difficult.
When it comes to support, we can consolidate products from a single vendor when there is no virtualization. But for storage virtualization implementations, the storage systems are of different models and procured from various vendors. Hence, the virtual solution needs to be well evaluated so that it addresses interoperability issues for storage virtualization success.
The whole storage environment is in control of the storage virtualization solution whose failure will lead to downtime of SAN infrastructure—causing inaccessibility of storage to hosts. So the storage virtualization solution should be fault-tolerant and robust with five nines of availability (99.999).
Different storage systems are capable of handling a predefined work load for storage virtualization. Hence the solution should be designed in such a way that the resources are utilized optimally, and that the applications are not starved of resources. Further, for storage virtualization efficiency, it's critical that the solution does not become a bottleneck and hamper application performance.
• Monitoring and reporting
Most solutions for storage virtualization available in the market lack monitoring and reporting capabilities. So for effective storage virtualization, the solution should integrate with existing monitoring and reporting infrastructure. This in turn will ensure that resource and storage utilization can be regularly monitored and reported to analyze the pattern of business activity.
About the author: Anuj Sharma is an EMC Certified and NetApp accredited professional. Sharma has experience in handling implementation projects related to SAN, NAS and BURA. He also has to his credit, several research papers published globally on SAN and BURA technologies.
This was first published in August 2010