The case for network-attached storage

Hyper-efficiency is a prime objective in e-business. Companies want to gain a competitive advantage by defining new sales opportunities faster than their rivals, fulfilling customers information queries quickly, and cutting their time to market with new products and services.

That often means finding a high-speed mechanism for filing and retrieving business data.

Network-attached storage (NAS) is one way around the slowdowns and service interruptions that occur in a conventional file server model. The shortcomings with the file server model were discussed in last week's newsletter, but essentially they center on the file server's role as a gatekeeper. In this setup, the file server sits in front of storage systems typically connected to disk arrays by a SCSI bus in a parallel cabling scheme. File servers read and write all storage requests from users and other servers, and only one file server has access to the disks that hold stored files. The file server can become a single point of failure or a bottleneck.

NAS effectively moves storage out from behind the server and puts it directly on the transport network. Unlike file servers that have SCSI and LAN adapters, a NAS appliance uses a network interface card to transmit LAN and storage communications.

These specialized storage devices bring with them some major benefits over file server systems. Because they don't require a file server, any user with access rights, anywhere on the network, can directly access stored data. This removes any delay posed by slow file server hardware or operating systems. NAS devices are also optimized to process storage I/O transactions much faster than conventional parallel SCSI schemes.

NAS appliances use a common file server access protocol, so they can process requests from systems around a network, running multiple types of operating systems. Any system, whether it is a Windows NT server or a Unix box, can access the NAS device. This solves cross-platform issues that arise frequently in today's heterogeneous enterprises.

NAS also mitigates the physical complexity associated with file servers that use parallel SCSI buses to connect servers to storage disks. NAS alleviates the awkward setups of parallel cabling, host bus adapters, and termination that are part and parcel of a file server implementation. Companies simply plug the system into the network and it is ready to go. Many NAS appliance vendors claim an installation takes less than an hour.

Yet for all its advantages over a file server model, one drawback of NAS is that it shifts storage transactions from parallel SCSI connections to the production network. This means the LAN has to handle both normal end-user traffic and storage disk requests, including backup operations.

Thus, NAS doesn't resolve one of the core issues associated with file servers - bandwidth consumption during backups. Unlike storage-area networks, which remove back-up traffic from the LAN, NAS still uses the production network for backup and recovery.

That said, NAS technology still supplies companies with an easy-to-deploy solution for resolving cross-platform issues while speeding access to stored files.

Amy Larsen DeCarlo is an analyst with Enterprise Management Associates in Boulder, Colo., a leading analyst and market research firm focusing exclusively on all aspects of enterprise management. She focuses on storage management, application management, and security. In her position, she oversees market research and contributes to custom project work in her focal coverage areas. Prior to joining EMA, Amy spent five years covering enterprise management for industry trade magazines, including InformationWeek and Data Communications. She can be reached at decarlo@enterprisemanagement.com.

Network Appliance

Auspex Systems

Dell

Archive of Network World on Storage in the Enterprise newsletters