Medium- and large-sized companies that take their storage seriously and want to streamline the process of accessing and backing up files may want to consider implementing a storage area network (SAN). SANs offer a shared back-up environment and give administrators the ability to consolidate their storage, achieve high availability, replicate data and better manage their storage.
At its most basic level, a SAN is a storage device that uses a Fibre Channel as an interconnect between the host server and the storage, but it’s also a combination of hardware and software required to manage the SAN itself. While many vendors are trying to sell customers on the idea of buying SAN equipment to extol the ease of implementation, Kanata, Ont.-based Kanatek Technologies Inc.’s Jim Pond, manager of professional services for Canada, explained that SANs are not plug-and-play technologies.
“That’s kind of one of the dirty little secrets about SANs -that it’s not entirely plug-and-play today,” said Pond, whose company’s core competency is architecting, testing and deploying SANs for its customers.
Defining the wares
According to Terry Retter, director of strategic technology services at PricewaterhouseCoopers in Menlo Park, Calif., SANs improve the flexibility of storage and are useful to companies with high-volume and high-demand server sites where reliability is important. A SAN can balance the workload of servers. SANs are often being employed in e-commerce, customer relationship management and enterprise resource planning environments.
The main hardware components for a SAN are a Fibre Channel switch, Fibre Channel host BUS adapters and Fibre Channel-enabled storage devices. Sounds simple enough, doesn’t it? Not really.
“That key part of it (hardware) is pretty well-defined. There’s not a lot of controversy about it,” Pond said. “The part that is not all that well-defined is what software you use -if you use any software -to actually enable that infrastructure. That’s where most of the effort is right now amongst the vendors -is to bring out value-added software to turn on new features and benefits of that snapshot of hardware.”
Software has been unsophisticated, and vendors are just now making more complex SAN management software, he explained. Pond said part of the problem is that there has not been an industry consortium to create standards for SAN technology, although the Storage Networking Industry Association (SNIA) of Mountain View, Calif., is currently pushing for adoption of SAN standards. Because of the lack of standards, many individual pieces of a SAN are not compatible with pieces from other vendors, he said.
Once the decision has been made to accept SAN technology into a corporate network, the company has to decide what it is trying to accomplish in terms of storage and what the company requires, said Wayne Hogan, storage sales specialist at Sun Microsystems of Canada Inc. in Markham, Ont.
Network managers and administrators should consider that the only real downside to SANs is that they create an added level of complexity that needs to be managed. Pond also noted SANs have a very high entry cost, although he said the return on investment (ROI) is worth it.
Selecting the equipment
After a company has figured out how much storage it needs and what its goals are for the SAN, it’s time to decide what brands of equipment it is going to purchase. According to Pond, there are approximately half a dozen SAN technology manufacturers. There are also vendors who have slapped their brand names on products to consider -original equipment manufacturer (OEM) partners acting as resellers.
Pond suggested two approaches to choosing equipment that will work properly together, keeping in mind the lack of proper standards in the industry. The first approach is to choose one of the OEMs, such as Houston, Tex.-based Compaq Computer Corp., and go entirely with their solution. It will be guaranteed to work because the vendor has chosen specific products from the various manufacturers to sell together as a SAN solution. The second approach is to contract a SAN integrator, such as Kanatek, to choose products and implement the SAN.
Assuming a company chooses to take the do-it-yourself approach, there are a number of things to keep in mind when going through the paces of SAN design and construction. According to Paul Ellis, director of marketing for storage at Austin, Tex.-based Tivoli Systems Inc., the first step to setting up a SAN is the installation of the hardware.
“You need to install the hardware and make sure that you can bring the hardware up as individual components and it functions as it is supposed to from a hardware standpoint,” Ellis said, “and then establish the physical interconnects. And frankly, all of that’s pretty easy when you get right down to it because it’s almost like taking a storage subsystem out of a box and plugging it into a server.”
He added that it is an easy task, assuming the person doing it is qualified for the installation.
Once the hardware has been established, it then needs to be determined which servers and applications will be allowed access to the storage, Ellis said. Hopefully this will have been planned out a little in advance. The storage will be divided up into logical units and then the logical units have to be attached to a particular application or server.