Enterprise storage users face challenges that go beyond simply looking for the lowest-cost tactical hardware procurement. One of the biggest challenges they face is to reduce their storage total cost of ownership (TCO) and manage storage growth more efficiently.
Storage area networks
Zophar Sante, vice president of market development for SANRAD, a San Francisco-based company specializing in IP storage networks, says one of his top recommendations for improving TCO and managing storage more efficiently is networking storage into a SAN to provide for storage resource sharing.
Networked Storage The Way To Go
One of the advantages of a SAN, says Sante, is that information lifecycle management (ILM) becomes an option. ILM is a great way to lower costs by moving infrequently used files off frequently used storage to a subsystem. “The data is still online, but the storage costs could be decreased from one-quarter to one-half the costs of the primary storage being used for frequently accessed files,” he says.
Scott McLeod, director of professional services at MaXXan, a San Jose, Calif.-based technology company offering storage solutions with scalable networks and integrated applications, agrees that ILM is an effective way to lower costs by moving infrequently used files off frequently used storage. McLeod also suggests designing and deploying IT Infrastructure Library (ITIL) compliant management processes (capacity management, change management, and configuration management).
Other industry experts suggest using SANs to centralize storage management. Evaluate what is being stored and don’t assume every data item requires the same level of storage services, they say.
“Look at ‘just-in-time’ provisioning to address storage use in smaller time intervals, rather than rely on multi-year predictions,” says John Lallier, vice president of technology at FalconStor, a Melville, N.Y.-based company developing network storage infrastructure software solutions. Lallier also suggests employing and implementing virtualization and storage services, such as migration and snapshots, on a set of disk vendor-neutral appliances to avoid hardware vendor lock-in.
— Peter Hunter, EqualLogic |
“Managing storage and reducing administration costs are at the center of an increasingly complex IT problem that challenges both small and large organizations,” says Peter Hunter, product marketing manager at EqualLogic, a Nashua, N.H.-based company that offers iSCSI-based SAN solutions.
“Today, the majority of storage is directly attached to servers, which is expensive, difficult to grow, and limited in management capabilities, especially as servers proliferate,” he says. Hunter believes that the migration of data to SANs is the cornerstone of all storage best practices. He says that storage consolidation provides the infrastructure for enterprise features and allows organizations to centralize storage management.
Experts agree that storage needs to be “pooled” so it can be allocated at will. “Volume virtualization, or logical partitioning, provides granular storage allocation, allowing storage administrators to allocate and resize capacity for a specific server and application,” says Sante. He adds that storage should scale independent of the network, hosts, and control layer so that the pool of available capacity can increase for the lowest possible cost (drive and enclosure only) without having to invest more in control software, network equipment or host agents.
Page 2: Classifying Data, SLAs Show Promise
Classifying Data, SLAs Show Promise
Eran Farajun, executive vice president at ASIGRA, an Ontario, Canada-based company offering online backup and restore solutions for network computing, says one of the things end users can do to better their TCO is to begin to classify data based on two simple criteria: data is either “critical” (needed for day-to-day operations) or “important” (required to for legal, compliance or other reasons).
— Eran Farajun, ASIGRA |
Once you start this process, you can begin to manage the data of the two classifications differently. “Simply put, end users that manage five-year-old data in the same manner as five-day-old data are doing a disservice to themselves,” says Farajun.
Farajun also suggests that end users take a high-level look at the architecture being used by various storage solutions (primary storage, backup and recovery, long-term archiving storage). Enterprises need to ask themselves if they are using outdated architecture for any of these storage operations. The market has new solutions to old problems, Farajun says. For example there are new methods of doing backup and recovery that might not have existed when an existing solution was purchased.
Farajun also suggests that end users attempt to introduce internal Service Level Agreements to their storage customers and consumers.
“When end users are able to set internal SLAs up, they will have an option to deliver specific storage operations as a utility across their organization, and therefore charge … based on consumption,” Farajun notes. If users know they will be charged for storage consumption, they will be much more careful about what they store, and the IT costs will be spread across the organization and not be incurred solely by the IT departmental budget, says Farajun.
IP Storage, High-End Features Offer Hope
IP storage is another way to do more with less by reducing or eliminating redundant networks and letting customers leverage their existing IP infrastructure and investments.
“Storage vendors are beginning to ship storage systems with native Ethernet connections, thereby negating the need for specialized FCIP networks, which provides huge capex and opex savings for IT end users,” says Robert MacIntyre, vice president of business development and marketing at Netex, a Maple Grove, Minn.-based data transport company.
The benefit of IP storage, MacIntyre says, is that end users can use existing IP infrastructures for storage networking applications. By reducing extraneous network and resource costs, customers will be able to divert more funds to storage systems, which in turn will help manage storage growth, he says.
Another suggestion is for end users to purchase enterprise-class features. Dramatic operational improvements in efficiency, reliability, and data protection have been developed in the form of enterprise features such as snapshots, data replication, RAID, redundant hot-swappable hardware and storage virtualization, says EqualLogic’s Hunter. These features are rapidly becoming more and more affordable and accessible to the broader market, and form the heart of today’s storage best practices, he says.
Hunter also suggests that end users automate their storage. Features such as storage virtualization and RAID are extremely complex and can create more problems than they solve in the form of administrator toil and overhead.
“Automation not only makes these technologies practical, it provides the intelligence to optimize their use,” says Hunter. “Combined with storage consolidation and enterprise features, automation is the glue that delivers the desired productivity improvements by streamlining operations.”
In the second part of this series on end user challenges, we’ll look at how you can improve your negotiating leverage with vendors when you come to the end of a lease or need additional capacity fast.