IBM Canada wants to drive intelligence into the network in a bid to bolster its storage virtualization strategy and address ongoing SAN limitations.
By moving the intelligence into the network, customers get access to a lot more information about the server environment (where the data is created)
and the physical assets (where the data is stored), said Brian Truskowski, general manager of storage software, IBM Systems Group in Markham, Ont.
“”Move key intelligence into the SAN itself to drive the benefits of SAN in the right direction,”” Truskowski said.
As part of the on-demand computing strategy, IBM released two storage virtualization solutions in June — TotalStorage SAN Volume Controller and SAN Integration Server — which are designed to address storage networking complexity today, said Truskowski.
While SANs offer a host of benefits to customers (including better connectivity, performance, scalability, distance flexibility, vendor and product choice and the ability to consolidate assets), Truskowski said there are still challenges hindering the full potential of SANs.
Customers can’t easily build and manage heterogeneous SANs because maximizing utilization of the physical assets is difficult. Multiple file systems cannot share data and must be managed separately, installation and configuration are problems and there’s no common way to view or manage the environment.
“”In the good old days, servers were attached to storage via a cable — an umbilical cord — and all of that storage was dedicated to that particular server. What we’ve done now is introduced the notion of a SAN and that’s opened up some unique benefits for customers, but there are still some things we need to do to allow customers to maximize the full benefit of a SAN.
“”SANs are good things, but we have work to do here.”” The path between a server and storage is still hard-wired — a legacy of the direct-attached storage days — and needs to be identified.
Servers and storage are still connected in a tightly coupled way — addresses in the server are connected directly to storage, he said. “”We haven’t really broken that umbilical cord yet, so there’s still a very tight relationship in most deployments of storage to server.””
Most of today’s technology is designed and deployed in “”this direct-attached world,”” he said. “”The functionality deployed up in the server has the ability to see a lot of different kinds of storage attached to the server, but it has very little knowledge of the other servers and applications in the environment. So although it has a good view of storage, it doesn’t understand any relationship with the other servers in the environment.””
Enter the benefits of storage virtualization. He said it alleviates the lack of understanding and isolation between devices, addresses interoperability challenges and helps users manage data.