With IBM, Upgrading Processing Capacity Is a Snap

Virtualization is enabling businesses—large and small—to stretch their IT infrastructure, budgets, and facilities farther than many business leaders dreamed possible, and all without sacrificing performance, reliability, or security.

But sooner or later, every successful organization hits the point where it has maximized the output of its infrastructure resources. At this point, any effort to accommodate growth with the existing resources risks degraded system performance, outages, major maintenance issues, and lost sleep.

In situations like this, capacity is no longer just an IT issue, it becomes a business issue. Additional server and/or storage capacity can't be postponed safely for very long. The typical IT department tries to postpone this day of reckoning knowing that such upgrades or expansions inevitably pose an entirely new set of challenges, some of which can't be foreseen with any degree of certainty.

Faced with this obstacle, every business owner and/or IT administrator wishes that they could just snap their fingers, and the capacity of the servers would suddenly increase by 50%.

Additional Power Is Easily Accessed
Well, it may not be a snap of your fingers, but IBM has made it almost that easy to quickly and easily expand your data center's capacity. IBM calls it Capacity on Demand (CoD) and has been offering it since 2000. In fact, owners of IBM's System i and System p may not realize it, but some of those systems already have that extra processing power built in and are ready to be activated on command.

What's particularly unique about CoD is that customers don't pay for the balance of additional processing power and memory until they need to use it. One way of looking at CoD is as a sort of "Virtual Capacity." As such, Capacity on Demand is entirely in keeping with IBM's brand promise of simplicity, flexibility, and efficiency.

The simplicity starts with the fact the IBM System i and System p are available with CoD as part of a customer's original purchase. Thus, the new buyer is prepared in advance for his or her first capacity upgrade, even though it may be one, two, or three years into the future.

The theme of simplicity is carried further by the ease with which the dormant processing power and memory can be switched on. When the customer decides that it is time to tap the dormant processor and memory, he or she places an order with IBM, and electronically encrypted activation keys to unlock the processor are sent by mail or via the internet. There is no hardware to ship or install, and no new contract is needed.

"What is particularly unique about CoD is that customers don't pay for the additional processing power and memory until they need to use it. One way of looking at CoD is as a sort of "Virtual Capacity." As such, it is entirely in keeping with IBM's brand promise of simplicity, flexibility, and efficiency."

Another inherent beauty of Capacity on Demand is its flexibility to suit the varying budget and processing needs of every organization. Clients pay for increased capacity only as their needs grow. For instance, the simplest situation involves a business that is growing and has a good idea of when it will require additional processing capacity. For these customers, the best solution is Capacity Upgrade on Demand. When the business grows to a point where capacity is becoming an issue, they can turn on the additional capacity and memory. They pay for this permanent expansion in capacity at that time, and can continue to grow without disruption.

Increased Capacity for Fluctuating Needs
Other businesses face greater fluctuations in demand for their processing. This varying demand for processing can occur because of seasonal activity, period-end requirements, or special promotions. They know they need additional processing and memory horsepower; they just don't need it all the time. For these businesses, IBM offers temporary capacity for fluctuating workloads. This offering is called On/Off Capacity on Demand. Users control when the additional processing power kicks in and when it is turned off, and they pay only for those periods.

In a further nod to customer convenience, there is Reserve Capacity on Demand, which provides automated use of on-demand processors for planned and unplanned short-term workloads that are run in a shared processor pool. The customer pre-pays for this capacity and isn't required to sign a contract.

When a customer needs temporary capacity for testing a new resource or application, Trial Capacity on Demand gives the customer the ability to evaluate how additional resources will affect existing workloads. In this case, a key is provided to activate the temporary capacity for 30 powered on days.

Finally, there is an option for the customer who will, at some point, require permanent access to advanced virtualization functions, such as logical partitioning, micro-partitioning, virtual I/O, or 5250 OLTP processing.

A further advantage of CoD (over redeploying a server that was mothballed following implementation of virtualization or buying a new server) is the ability to bring new capacity on line without all the disruption, installation, and prep time that a new piece of hardware can entail. With CoD, processors and memory can be activated dynamically without interrupting system or partition operations. In addition, processors can be activated in increments of a single processor, while memory can be added in one Gigabit steps. Again, this allows the customer to pay only for what is required to get the job done.

With Capacity on Demand in System i and System p, users need not worry that their data centers will run out of head room quickly, nor do they have to pay for processing and memory capacity that they might not use for years. IBM makes IT simple. Capacity on Demand is one more way customers can experience that simplicity.