After a few years out of the storage end of the business, one of my clients recently asked me for recommendations about which storage vendors to consider in an RFP process. After digging around for a few weeks, I came to a not entirely obscure realization: Despite their best efforts, storage vendors still fall into two distinct tiers—storage solution vendors and service providers. As the technology gap between the two tiers has narrowed, prices have not narrowed accordingly, except in instances of sales. This indicates that, despite their claims to the contrary, something—other than raw throughput and storage capacity—creates and maintains the division. Understanding that something allows us, as customers, to make more informed decisions about what companies we consider during an RFP or other purchasing process.
Storage solution vendors vs. service providers
A quick scan though the arrayed legions of storage vendors shows just how rapidly the technology gap has narrowed. High quality technology, ranging from propriety solutions to new iSCSI devices, constantly pushes the edge of what was possible "back when." Disk storage has finally become cheap enough to, in some applications, out-compete removable tape devices for archive purposes. Infinitely cheap storage seems right around the corner.
More importantly, the "threshold" of technology has reached the point where, except in extremely intense applications, even the cheapest solutions can meet the majority of a company's transaction and storage needs. Solutions have also become highly sophisticated in their interactions: With a little care, you can build systems exactly matched to your transaction/storage/recovery needs.
The vast majority of these products come from storage solution vendors. This first tier of providers works hard to produce innovative, quality technology solutions to a variety of everyday IT problems. They work within established or emerging standards to give people relatively interoperable tools ranging from NAS to tape libraries. The vendors and their affiliates also provide some "storage consulting," warranties, and usually on-site system repair if something goes wrong.
A small handful of products, almost universally the most expensive ones, boast marginally better technology. In technical presentations, they make much of small differences that will, for most customers, simply not make enough of a difference to justify the cost. However, where the real difference comes is in the services (equally expensive) and software tools attached to the product. By buying the product you gain the ability to pay them for these additional services. If you have terabytes of mission critical data with millions of transactions, these expensive services mean the difference between life and death; the other systems might handle the load during normal operations, but when something goes wrong (and it will…this is IT) the downtime can bring a business to its knees.
These second tier service providers monitor your system in real time, respond to problems before they develop, and are right there with you when something catastrophic occurs. The up front and continuing costs can seem steep, but not when compared with the cost per minute in lives or in money accrued by the systems that need their help.
Choosing which is right for you
Which of the two types of storage providers you should consider depends entirely on a realistic assessment of your specific situation. Before you begin the RFP process, answer the following questions realistically:
- Do you have more than one storage system? If so, does it need to be consolidated?
- What happens when the storage system goes down now? If you are consolidating, what will happen when you move to the new system? What calculated risks will the company run if one or more of the systems fail?
- Do you have the capacity to manage the system for your uptime requirement yourself?
- Do you have the capacity to restore service in a reasonable amount of time?
- What other processes or systems will the installation of the new storage system affect?
The first and second question-sets assess the undertaking's scope. In terms of vendors, a small scope with low risk (e.g., the consolidation of several low-risk file servers) can easily be met by a storage solution. A large scope with high risk (e.g., the consolidation of several active business-critical databases and their development platforms) calls for the kinds of mitigation techniques available though the more expensive service providers.
The third and fourth question-sets assess your own internal capabilities. If you have a capable staff and a solid, tested disaster recovery strategy, you can afford to go with less comprehensive services and potentially less user-friendly software management tools. If you have large amounts of critical data but limited ability to manage it and the risks associated with it, a service provider can give you the risk buffer you need to develop those capabilities.
The fifth question leads us into dangerous territory. We think of storage as a basic service. Our users, however, see storage as both a given right and a limitation on their activities. Each process that uses the storage system wants as much storage as it can get with as little possible interference on our part. Take, for example, ERP development: I know of shops where they manage four separate instances of the live data in order to ensure quality control, each instance absorbing roughly 500 GB of storage space. Could they get away with less? Possibly. Could they reduce the number of instances now that the development team has become used to it? Unlikely, unless they developed more political clout and the ability to override the people with organizational responsibility for the company's data transactions.
The more complicated the answer to the fifth question becomes, the more we, as infrastructure managers, start to look at service providers over solutions providers. Service providers may cost considerably more but the reduced (for us) maintenance times and more sophisticated software tools mean we remove a source of contention in the environment—that in turn allows us to focus on more important activities.
Application of the questions: One client's approach
One of my clients used this set of questions when considering whether or not to consolidate its laboratory data storage systems. Unfortunately, they realized several of their products with extremely long (12+ months) production processes required nearly constant monitoring to ensure successful manufacture. The rest of their data (about 800 GB of assorted files and mail systems) could go down for up to 24 hours without meaningful impact on the business.
Running though the questions they found that they really needed two separate storage solutions: one ultra-high availability using a service provider and another using low-cost, solution-provider equipment. They then took the idea one step further: Rather than complicating their environment, they approached the service provider with their information. As a result, the service provider cut its prices in order to make the ROI on its service better than that of the dual solution. In the end, my client got far more than he needed but at a much more palatable price than he thought possible.