IBM sees virtualisation as next step in evolution of data management
WITH MASSIVE amounts of storage on the horizon for every business that generates digital information, simplifying the way it is handled will become extremely important, according to IBM.
Kelvin Li, zSeries manager for storage at IBM in Hong Kong, said the key to this would be infrastructure simplification.
'Viewed in the broadest sense, infrastructure simplification represents an optimised view and evolutionary approach, or the next logical step beyond basic server consolidation,' Mr Li said.
'From an IT perspective, infrastructure simplification leverages existing assets, enables business process integration, reduces costs, increases efficiency and moves to better synchronise the design of IT infrastructure with the design of business process.'
Mr Li said it was a critical part of business today that applications stayed up and running, and that people had access to the data being created, shared and stored.
IBM's Information Lifecycle Management (ILM) and virtualisation technologies were created specifically to make the management of all this information easier.
'ILM is a process for managing information through its life cycle - from conception to disposal - in a manner that optimises storage and access at the lowest cost,' Mr Li said.
Because of its long history, IBM was uniquely suited to handle this problem, he said.
One area where IBM hoped to lead the industry was storage virtualisation.
'Storage virtualisation can provide a single logical view of all the physical storage resources within a storage network, so that available capacity on disparate storage devices can be viewed, managed and provisioned as a single entity.'
This kind of virtualisation will become an integral part of IBM's product offerings.
'SANs (storage area networks) have helped control the booming growth of storage by detaching direct-attach storage devices from individual applications and connecting them in a switching environment to enable the physical pooling of storage and servers,' he said.
This physical pooling can help, but is not enough if one wishes to get the most out of a system.
'The popularity, diversity and size of SANs have presented IT managers with new issues of manageability, such as increasing numbers of pooled arrays and a growing variety of storage controllers.
'Block-level storage virtualisation stands on the threshold of maturing as a viable commercial solution for tackling the growing issue of complexity in data storage management,' he said.
At the file level, customers within their data centres and their SANs often have dozens, or even hundreds, of servers. These servers may well use different file systems, and this means that although the storage may be pooled together in a SAN, the different host applications are unable to share data because of the different name spaces and conventions between the file systems.
This is easily handled with the right kind of virtualisation.
'File-level storage virtualisation will enable a unified name space to allow file sharing across multiple heterogeneous or homogeneous host environments,' he said.
