Fabric virtualization: the roadmap: Visualize, virtualize, automate. (Business of Technology).

Envision an entire IT infrastructure that acts as one logical network, managed through a single point of control. This "smart" network fabric will readily support and adapt to changes in the business environment, improving customer relations, increasing productivity, ensuring business continuance, and helping to contain costs.

Ten years ago, an IT infrastructure largely meant hardware. Not so any more. Today, the applications--ecommerce, ERP, CRM, email, etc.--are the driving force behind a growing IT infrastructure. But the infrastructure applications we currently rely on are evolving. Only a few years ago, they were products that were either installed individually on each PC for personal productivity tools, or on a mainframe for business applications. These systems tended to lock information in silos and were expensive to manage. Customers today demand improved ROI from their IT infrastructure. They need solutions.

Centralization of servers and storage is the roadmap to improved ROI. Group planning and email, for example, are accessed at the desktop, but they reside in a centralized data center exchange server, saving the company management costs while improving the utilization of storage resources and increasing productivity of the business. The end-user never sees the product, but has all the access he or she needs to the service.

Storage: A Growing Expenditure

Today, with the proliferation of data-hungry applications, storage has become the center of the IT infrastructure, and networked storage is a necessity for many (if not most) businesses. Even as the economy has slowed, storage capacity is still growing at exponential rates. IT organizations are faced with the daunting task of integrating new applications and bringing all of their current business systems up to date, while, at the same time, keeping storage costs under control. According to IDC, the number-one-deployed external storage type will be storage area networks (SANs). This will bring significant benefits to the IT organizations, as well as new challenges.

At the heart of the SAN is the storage network, and as Meltzcafe's Law implies, the value of the storage network dramatically increases as the number of servers and storage systems are networked together increase.

IT Challenges

To support the data demands of mission-critical applications, the IT organization faces many new requirements. It is the job of the IT department to deploy new applications and the supporting storage infrastructure in a way that satisfies the following:

* Optimize existing technologies, keep pace with a rapidly growing information base, demonstrate the economics of the infrastructure, keep track of a myriad of new technologies, and still keep best practices in mind.

* Deliver technology that meets the requirements of all of the company's applications at a rapid return on investment and improved total cost of ownership.

* Minimize the risk and complexity of the solution they are implementing. At the same time, they need to speed up project implementation so the enterprise can take advantage of their benefits more quickly.

* As they deploy new solutions, it is imperative that IT executives also protect the investment they've already made to the infrastructure.

And, perhaps most importantly, they need to build a reliable, scalable, highly available IT infrastructure that efficiently leverages its limited re-sources. Bottom line, it needs to manage more with less budget dollars and time, keep their valuable IT resources focused on strategic business-critical initiatives, and improve the utilization of the complete infrastructure, while still enabling its company to increase productivity, give it a competitive edge, and increase the performance of the applications it relies on.

Not an easy task. On the horizon, there is a cost-effective, scalable solution that can answer this problem.

Fabric Virtualization 

As storage evolves, the intelligence moves into the network itself. The system becomes centrally managed, but functionally distributed. Every device on the intelligent network will be application-aware, a specialist in its own data storage or trafficking function, turning the fabric itself from a collection of nodes, switches, servers, cables, and software to one transparent service for the end user to access. McData calls this transformation the network into a service that creates value: fabric virtualization.

Virtualized fabrics are network-focused and will be able to configure themselves, discover and diagnose problems, and heal themselves without operator intervention. Fully integrated into the fabric of the storage network, future applications will give customers comprehensive automated management tools that operate invisibly from a single point of control in multi-protocol, multi-vendor environments.

A good analogy is the phone system. Anyone can pick up their phone, call across the world and get a quick connection at an established cost. Now think about what goes into making that call. Do you know which infrastructure supports your call? Was it SONNET, VoIP, ATM, etc.? How was the call routed? Did a device fail, and the system seamlessly reroute your call? You probably will not know the details and complexities that needed to happen to make the call a success. And you don't really care, do you?

Just like the phone system, the storage network of the future will actually be a solution that acts like a service, connecting servers to storage. We expect:

* Low latency.

* A clear, uninterrupted transfer--in other words, high-quality service.

* A secure transaction and secure data.

* A predetermined cost, which has been established to support a service level agreement.

* And, finally, an invisible infrastructure. The end user doesn't care about the underlying protocol or how the information is routed. The end user cares that the service is provided quickly and effectively.

Fabric virtualization evolves the storage network into a true service that connects servers to storage. Intelligence is built into the fabric, allowing the fabric to respond to demands from applications that are at the heart of the enterprise.

Ultimately, through route management and self-provisioning, the focus of storage management itself will move away from the individual devices on the network and into the fabric itself. The physical attributes of the infrastructure will become more transparent and delivered as features such as quality of service, performance, availability, and security, all managed in a simple, logical abstraction.

IT Implications: A Real-World Scenario

What does an invisible fabric mean to IT personnel? Optimal utilization with minimal system administration.

Suppose your company makes computer games. During peak season, you've come out with a game that you're selling on your ecommerce site. Your marketing department runs a special promotion, and suddenly you're getting so many hits, you can't handle the traffic. You're losing sales every minute, because your ecommerce system handles only a certain number of calls per hour, and demand is far outpacing what it was designed for. Sound familiar?

Application aware systems that leverage technologies such as fabric virtualization will change all that. The IT manager can prioritize and allocate bandwidth so that the system delivers the best possible performance to mission-critical applications at crucial times of the day. And they can determine optimal routes for the data to travel, and build in a redistribution plan in the event there's a failure in the fabric.

A virtualized fabric will dynamically allocate resources according to pre-determined priority settings. From 6 a.m. through 10 p.m., your game sales on the Web may take priority over data backnp. At those times, you can schedule all available ports and routes to be allocated first to the Web server. Late at night, when backup is the highest priority and traffic is very low, the Web server can take a back seat to other processes. The virtual fabric knows that and routes incoming data as necessary.

The roadmap for fabric virtualization will emerge in three distinct phases: visualize, virtualize, and automate. Your organization may already have monitoring software to help keep track of devices in the storage network--useful but not nearly enough to meet increasing data availability, scalability, and security needs. Look closely at your needs and your growth patterns, and measure your system--visualize where you need more bandwidth, more capacity or security. Applications such as SANavigator can help manage all the devices on your storage network by aggregating them into a fabric level of management that will save an enormous amount of time and increasing productivity.

Next, virtualize and automate your storage environment to optimize ROI and reduce total cost of ownership. The more advanced applications of the future will take this one step further, proactively managing the network and calling for service before there's a degradation in the service that the fabric delivers.

IT executives need to have a vision and a strategy for a storage network that is scalable, flexible and highly available. The infrastructure that you start deploying today must be able to support this roadmap, otherwise, you may have to phase them out when your requirements grow.

Faced with exponentially expanding data, limited resources, multi-vendor and multi-protocol environments, and cost-containment issues--the yellow brick road for IT managers leads straight to storage networking.

Brandon Hoff is senior manager of strategic marketing for McData (Broomfield, CO).

www.mcdata.com