The total cost of owning and managing storage varies greatly from company to company. IT administrators must weigh a number of factors such as customer needs and business applications when designing and implementing storage environments. An IT department that understands the hidden costs of unmanaged storage will most likely end up with a lower storage TCO.
As storage technologies continue to emerge to meet the demands of mission-critical applications, some IT professionals are asking how organizations can avoid the hidden costs associated with unmanaged storage.
You first have to gain a sound understanding of the business value of your data and what is essential to your business livelihood, says Mark Eastman, product marketing manager at Quantum DLTG. “By understanding what data is critical and by having a solid storage management plan, the threat of being impacted by hidden costs is minimal,” he says.
Jeff Hornung, vice president of business development at Spinnaker Networks, believes that one of the ways companies can avoid the costs of unmanaged storage is by driving toward the best in class IT staffing levels and through better implementation of SAN and NAS.
As obvious as it may seem, the key to avoiding the costs of unmanaged storage is creating and implementing a carefully thought out management strategy, says Jon Greene, director of product management at FalconStor. Companies must address the following questions:
- What resources need to be managed?
- What are the business requirements for data availability, analysis, protection, regulatory compliance, etc.?
Greene believes that by carefully mapping out these requirements, enterprises can determine the appropriate storage topology and management strategy. He also says that the decision is not simply, “Do I implement a SAN or not?” Instead, the decision needs to take into account the type of solutions to be used and must also address the following issues:
- Are the decentralized storage services (virtualization, replication, etc.) at hosts or storage arrays sufficient?
- Will they provide the required control and configuration flexibility, or do I need centralized management and services?
- If I need centralized management, should it be passive (i.e. a storage resource management reporting and planning solution) or active (a SAN storage service solution to provide SAN/NAS provisioning, security, high availability, and data protection/disaster/recovery services)?
Total Cost vs. Capacity
Storage unit costs invariably decrease with capacity. While the total costs for DAS and SDAS are initially less expensive than SAN for a certain level of storage capacity, some industry analysts say that enterprises must realize that total costs for DAS/SDAS increase at a much faster rate than for SANs.
“This premise seems to contradict all public studies which I have seen as well as conventional wisdom,” says Ken Steinhardt, director of technology analysis, EMC Corp. “The greatest component of TCO that makes DAS expensive is the cost of staff resources, which will continue to rise as hardware costs fall.”
Steinhardt also says that shared DAS can in fact create additional cost issues beyond burdening important staff resources.“Customers buy CPUs to run applications, not to be burdened by remote storage requests from someone else’s CPU — which is always the case with shared DAS,” he says.
Eastman agrees from the hardware and software cost perspective, but not from the management of costs view. “Any TCO evaluation needs to address all three aspects,” he says.
Generally, the larger the capacity, the more the SAN infrastructure can be spread across the terabytes of storage, maintains Hornung. “As capacities increase, acquisition costs per gigabyte for SAN/NAS will decrease, while the opposite happens for DAS — the cost per gigabyte to manage goes up as the number of individual DAS systems increases,” he explains.
An Issue of Understanding Storage TCO?
As storage issues continue to be batted around by industry analysts, another key point is the question of whether or not enterprises in general actually understand the composition and drivers of storage costs. “That’s an excellent question,” says Steinhardt, “as while it is impossible to generalize, there tends to be a grey area with regard to this issue; most enterprises either understand these issues well, or they do not.”
He maintains that those who do not have a good understanding of storage TCO tend to focus on hardware acquisition costs, often at the far greater expense of the other components that drive TCO. “These other aspects of [storage] TCO could include the efforts associated with configuration, monitoring, provisioning, back-up, data-replication, integration with third-party software, multi-vendor integration and pre-testing, reporting, the ability to adapt to new technologies, and more,” says Steinhardt.
Green believes that because storage has been the focus of so much IT attention in recent years, and with the intense pressure on budgets, most enterprises have a much greater understanding of their storage cost structure. “Enterprises have recognized for years that administrative and operational costs dwarf acquisition costs over a storage asset’s lifecycle,” he says. “Downtime costs are application dependent, but in many cases a small outage is catastrophic.”
“Unfortunately,” he continues, “until recently, businesses have lacked an integrated solution for reducing those costs.” Green reports that in the last few years comprehensive storage management services and solutions have established themselves in the marketplace and, when used in conjunction with SANs, have been providing cost relief for larger enterprises. “Increasing volumes for traditional CSAN technologies, combined with emerging standards such as iSCSI, holds the promise for providing similar relief to the mid-range companies,” he concludes.
Hornung does not believe that enterprises understand the composition and drivers for storage costs. “But,” he says, “they are getting better. The hidden costs are not the costs of acquisition, but IT management and client/application downtime associated with expansion and reconfiguration.” Hornung’s advice to companies: “Take a close look at those costs in your company.”
Eastman said he recently had a discussion with an IT manager of a major university in Chicago. Eastman asked the manager about his backup strategy and the gentleman laughed. “He knew the costs for additional disk and tape, but simply stated that he did not perform backups but once a year because he did not have the resources,” says Eastman. “I have seen many formulas for TCO, but my simple suggestion to companies would be to roll-up the costs as an investment in the future, because money spent today will offset management costs in the future.”
A critical piece of storage infrastructure planning revolves around optimizing a company’s storage resources. Poor planning can mean disruptions because of unplanned service needs, hardware addition or removal, and other non-application-related operations. Companies that understand the composition and drivers for storage costs and spend money to carefully plan an optimal storage environment can avoid downtime and thereby save money over time.