Through 2003/04, infrastructure consolidation will be driven by value-based portfolio management, but remain impaired by non-linear server pricing, immature tools, service-level priorities, chargeback, and organizational politics. Physical co-location and networked storage consolidation will be widespread during 2002/03. Premium high-end server pricing, coupled with immature partitioning and workload management, will hinder higher-level OS, DBMS, and application server consolidation for Unix (until 2003/04) and Windows (until 2005/06).
We believe 2003 will continue to be a difficult year for ITOs, as overall IT budgets remain flat to slightly increasing. We also believe storage will continue to increase its portion of the overall budget (most likely at the expense of server hardware). Indeed, storage commonly constitutes 60%+ of the hardware budget and can constitute 15%+ of the total IT budget when people, hardware, software (both disk and backup/recovery), and maintenance are included. Whether storage is 12% or 18% of the IT budget varies among organizations, but many CIOs and CFOs have realized, sometimes alarmingly so, that the percentages have steadily risen during the past five years.
By 2007, most ITOs will have formal policies associated with storage procurement, management, usage, operations, etc. Through 2007, ITOs will continue to struggle with tactical financial decisions of whether to purchase a particular storage hardware, software, or backup/recovery solution. Indeed, this is one of the challenges facing the storage resource management market, because immediate return on the initial investment is not always quantifiable.
Despite these tactical challenges, disk storage procurements must be made as data center storage capacities are still increasing 40%-60% annually, and though hardware prices are declining approximately 35% annually, more money will typically be spent on storage hardware. Here, we discuss three easily identifiable cost justifications for a disk storage purchase. This exercise must not be done in a vacuum and – particularly for data center implementations – strategic focus (with regard to the technology and vendor) must be paramount.
Many enterprise storage subsystems (e.g., EMC, HDS) purchased during the 1999/2000 “dot boom” and subsequent “dot bomb” are approaching the expiration of the typical 36-month hardware warranty period. Frequently, it is less expensive for an ITO to replace 36-month-old storage hardware (typically the useful life anyway) with a current technology than it is to begin maintenance payments. Although disk capacities have probably been added to an existing subsystem/frame, the warranty on the physical disk drives (as well as any cache or other hardware added after initial acquisition) is co-terminus with the initial subsystem implemented. Despite not being the preferred warranty policy, it is common (and accepted) due to vendors’ inability (or decision not) to track multi-initiated warranty periods for large quantities of hardware components.
Maintenance pricing can be as high as US$25-$30/GB per year, which makes the justification for extending the useful life of storage hardware for 48-60 months extremely difficult. Seeing this as one of the cost-saving components, ITOs can decide to trade in existing storage hardware and use this in the justification for storage rationalization (reduction in the number of types of things). Yet, depending on storage class, some ITOs may find a useful life for the storage in test and development or other “non-critical” systems. If this path is chosen, ITOs should examine a “do you feel lucky” strategy and not pay maintenance, and simply employ a time and material maintenance contract.
Given the economic climate and overall expected IT budgets in 2003, this is a buyer’s market. Although good for ITOs, storage vendors, as they did in 2002, must make concessions. Concessions were not commonly seen in the past but are more common to win the business and help economically justify a storage consolidation/replacement strategy. Our research indicates buyout offers vary widely depending on several factors: customer profile, the desire to win the business, the potential for additional near-term business, future longer-term business, storage capacity, storage class, useful life of the storage that is being replaced, and the total price tag of the proposal.
Some vendors do not make a practice of carrying other vendors’ inventory, but based on the desire for market share and mind share awareness, this is something ITOs should insist on to help justify a purchase (after any depreciation or final lease payments are incorporated). The purchase price of the new subsystem/solution, the trade-in value, and the financing/lease rates should all be negotiated individually but take into consideration the total “drive away” price.
Return on Investment/Total Cost of Ownership
ROI and TCO are metrics that senior management (e.g., CIOs) has increasingly been asking for during the past 24 months. We believe what senior management really wants, when inquiring about particular consolidation plans, is to know how quickly results will be seen and what those results will be. Although there are benefits to consolidation plans that can be quantified, there are many that cannot. We recommend clearly identifying the “quantifiable” and separate the “squishy” benefits. Clearly, specific monetary savings can be associated with high-profile applications where revenue per minute/hour metrics are known, but there are others, such as operational, infrastructure, comfort, recoverability, and other factors, that may be improved on but may not be able to broken into concrete monetary amounts.
ITOs looking to upgrade particular storage environments may be able to have quantifiable information with regard to particular application(s) that can be used to justify a purchase. Upgrading midrange storage with enterprise-class storage can yield benefits from hardware and software that contribute to an application performing a business function. ITOs should also ask for competitive trade-out discounts for software when changing vendors (particularly in the backup/recovery market).
However, in the disk storage market, this is a more difficult negotiation tactic. Qualities such as decreased recovery time for an application (using disk based replication capabilities), increasing capacity utilization (offset with declining hardware prices), increased performance (more transactions per minute), consistent management across more servers (more managed TBs per administrator), and better pricing with a strategic storage partner can provide immediate, measurable returns on the investment.
Although these metrics will not be an all-encompassing calculation for consolidation (which can be daunting), they may provide justification (along with maintenance and buy-out offerings) for a tactical purchase in difficult economic times. These decisions must be made as part of a strategy and plan for implementing an adaptive storage infrastructure.
Business Impact: Even in difficult economic times, looking at tactical storage purchases and cost-saving metrics can aid in more rapid application development and deployments.
Bottom Line: Short-term tactical justifications for storage purchase and upgrades can often exist. However, ITOs must not lose sight of the strategic importance of an adaptive storage infrastructure in the process and during difficult economic times.