Virtualization Technology News and Information
Storage Virtualization - Getting it right

Quoting from ComputerWorld

Data owners rarely classify data, making the data’s existence, value, and criticality frequently unknown to others. This creates compliance problems, with today’s data protection technologies typically addressing availability, but not confidentiality or integrity.

Speaking at a recent Computerworld IT Executive Summit on Simplifying Storage, Rajesh Prabhakaran (left), chairman, Storage Networking Industry Association (SNIA), South Asia, said that data security dependencies and problems are poorly understood by most enterprises and inadequately addressed in new technologies like virtualisation, grid computing and utility computing. Data aggregation and storage consolidation are creating potential failure points, he said.

With the information explosion, data is being created at an exponential growth rate of 60 per cent yearly. According to Prabhakaran, about 99 per cent of data is at-rest in storage resources while one per cent is in-flight.
As they struggle to cope with this growth, one of the largest challenges organisations face right now is articulating a plan that defines what they want their enterprise storage infrastructure to look like. “Without a vision, organisations generally resort to buying and managing storage based upon the two lowest common denominators: who has the cheapest storage and which vendor’s name is on the box,” said storage analyst Jerome Wendt.

While that may work for the short term, that is not a sustainable long-term model. Though price and reputation are worth something and should be a factor in the decision-making process, someone needs to look beyond the daily tasks of allocating storage and responding to crises and ask critical questions before just buying more storage. According to Wendt, questions like “What do I want my storage infrastructure to look like?” and “Will this purchase move me down that road?” are good starting points.

“Without knowing the answers to those two questions, you end up with the kludge many organisations try to manage now,” he said. “Disparate systems with no central management interface, different pockets of individuals with ­varying levels of expertise and no one who has the knowledge or authority to take companies down a path of change. Yet to establish a storage vision requires organisations to possess some type of baseline information so they can identify where they are at now before determining where to go next.”

Classification of data will enable more efficient storage utilisation and lay the foundation for consolidated operational and disaster recovery practices as well as archiving to meet compliance needs. “In the future, new data classification and management systems will be comprehensive, policy-based and application integrated as well as business driven,” said Prabhakaran.

As information grows, operation costs are getting out of control and out of proportion. Storage virtualisation is an area where companies are looking at to solve this problem. One of the stated purposes of virtualisation solutions is to create and maintain virtual volumes of data that applications can always access, regardless of the physical location of the data. “This capability enables IT managers to reduce wasted capacity through the use of oversubscription and enable non-disruptive migration of data to different storage systems,” said Prabhakaran. An inevitable consequence of volume virtualisation, however, is the obscuring of the link between the physical storage assets and the applications.
Effective capacity planning, performance management and root-cause analysis are some of the areas where virtualisation can make the management task more difficult for IT administrators using current tools. “When dealing with storage virtualisation, you need to have a management of change,” said Prabhakaran.

Companies selling storage virtualisation solutions must go beyond simply adding virtualisation products to existing storage device management systems. They must develop solutions that support new processes like storage pool management, non-disruptive data migration, and end-to-end performance monitoring. “Only with these additions will storage virtualisation successfully evolve from leading-edge or limited deployments into true enterprise-wide solutions,” Prabhakaran added.
Claude Lorenson (left), group product manager, Windows Server division, Microsoft, said the industry will need to keep providing features and functionality while driving simplicity.

There are no single-vendor data centres any more, and interoperability is vital for customers, he said. Microsoft’s Windows Storage Server, for example, aims to reduce storage and capacity and storage management overheads with Single Instance Storage (SIS). SIS replaces identical copies of files with links to one copy on a volume, in a way that is transparent to users and applications. This eliminates the need to track or condense duplicate files. It also saves backup tape space, with many backup applications already using SIS application programming interfaces, said Lorenson.

Read the original article, here.

Published Wednesday, June 07, 2006 6:19 AM by David Marshall
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<June 2006>