JUSTIFYING MAINFRAMES FINAL DRAFT – Bernell 1
1240 words
REFRAMING THE MAINFRAME
Submitted by Laura Bernell for BMC
August 26, 2007
MAINFRAME VALUE MAY TRUMP COST
IT Administrators who put business first are finding that measuring the worth of a mainframe by cost alone may be deceptive and counterproductive. In many cases, the value of a mainframe far exceeds its cost.
The value of IT resources is best evaluated through a business framework, even when an IT Administrator is the one doing the evaluating and making the decision. A business framework requires IT Administrators to examine the services that a given resource delivers to the business, and then analyze how those services increase the business’s productivity and profits. Making value-based decisions is leading IT Administrators to choose the mainframe in cases where the strength, capacity, and power serve business best.
For example, a server that supports online stock trading is of higher value to the business than a server that supports online access of employee-benefits information. IT Administrators who are also business-oriented realize in an instant that temporary loss of the online stock trading service could be catastrophic to the business, both in terms of bottom line and public image, while temporary loss of employee benefits information system has far less business impact. Therefore, they would choose to put the stock trading application on smaller, more nimble machines and the employee benefits information on the stable, more capacious mainframe.
A WORKHORSE, NOT A DINOSAUR
A mainframe is a workhorse, not a dinosaur. It can handle extremely large databases and huge transaction volume—and it can do so with unmatched reliability, availability and serviceability. A mainframe is like one of those impressively strong and muscular draft horses, descended from war horses, that pull plows and haul freight wagons.
Evolution of the Mainframe
Mainframes are evolving as robustly other components of IT. No longer serviceable as the stand-alone, isolated computer in some enormous and neglected data processing center, they have been adapted to the integrated, networked world. They now have mainframe software components that can support Web services in a service-oriented architectural environments. These Web service abilities permit the Mainframe to tap into the power and stability of legacy systems, such as CICS and IMC, in new Service Oriented Architecture applications (SOA), one of the inexorable forward-moving trends of modern IT. The mainframe is not going the way of the dinosaur.
Mainframe management tools continue to advance, too, enabling the mainframe to automate many tasks. This in turn has made it easier to manage and administer the mainframe. And self-diagnosis capabilities that simplify problem resolution have been developed. These advanced tools augment the skills of the operator, and sometimes make up for human error. These advances in mainframe tools also boost the productivity of the mainframe operations staff, so more mainframe applications and databases can be added to the data center without hiring more people.
LOAD CAPACITY,STRENGTH, RESILIENCE
The mainframe has unique features that render it stronger, more powerful, more resilient, and more able to sustain a heavy load than other platforms. These features support its strength, like the skeleton supports the muscle and heft of a magnificent horse.
Virtualization Technology- What you See is More Than What You’ve Got
One of these strength-supporting features is Virtualization technology. Virtualization provides the ability to run multiple server images — called virtual servers — on a single physical host server. Each virtual server appears to the environment as a separate physical server. Virtualization has been cited by Gartner as one of four key trends set to drive technology further into the business.[1]
Virtualization actually originated on, and is more highly evolved in, the mainframe. LPAR [DEFINE THIS] permits a single physical mainframe to host multiple mainframe images, with each image appearing as a separate operating environment. This in turn allows delivery of a wider, more fault-tolerant variety of services without increasing the number of mainframes in the data center.
Clustering –Nothing is Indispensable
Another strength-supporting feature of the mainframe is clustering; that is, the ability to [DEFINE]. This also originated on, and is more highly evolved in, the mainframe. Using Parallel Sysplex, the mainframe can integrate multiple, co-located physical mainframes into a cluster, with the cluster appearing as a single system image managed from a single point. This in turn enables failover to prevent service loss in the event of a system failure, allowing dynamic load balancing to maintain rapid performance, even under extremely heavy workloads. Furthermore, this capacity is extended with geographically dispersed parallel syspelx (GDPWS), so that business can continue without missing a beat, even if a disaster such as fire or flood takes out an entire data center!
THE RIGHT PLATFORM FOR THE RIGHT JOB
We all learned to use the right tool for the right job. Similarly, choosing the platform that fits the application, processes, and business-services for a given job is just good business. IT administrators are finding that when they need power, strength, and enormous data storage capacity, they look to the mainframe. If they need to avoid a hiccup of disruption or immediate response to a user query, they look to a smaller server platform. Cost is only one of several factors in consideration of a tool. You wouldn’t use a pinking shears to lop off a Ponderosa pine branch looming over your rooftop. And you wouldn’t use a chain saw to prune a rose bush.
IT Administrators are choosing their IT tools based on a clear understanding of each data center platform’s unique strengths and serviceability, and matching those strengths and capabilities to the job at hand. Simply transferring jobs en masse to lower cost platforms has been shown NOT to be the best formula for success. Administrators are balancing cost with business value. When they do this, they find that the mainframe is irreplaceable when it comes to certain applications and databases, and can readily justify its cost.
MAINFRAME HANDLERS
There has been a perception in the industry during the last decade that the pool of people with mainframe-related technical skills is drying up. This fear is ill-founded and not supported by the facts.
Interest in and training for careers in mainframe careers is alive and well. More people are opting for careers in the mainframe arena. Conversely, the mainframe industry is nurturing this revitalized interest with new training offerings in a variety of mainframe-related subjects. [NOTE: THIS SEEMS WEAK. DO YOU HAVE ANY SUPPORTING FACTS?]
OPTING FOR THE MAINFRAME: A REAL-LIFE CASE
With more than $2 billion in annual revenues and a market capitalization of $3.5 billion, Aliant is the leading information and communications company and the largest publicly traded company in Canada. The company employs 8,400 people and has two million residential customers and 80,000 business customers, including more than 170 enterprise customers that generate more than $500 million in annual revenue.
As the core platform for many of the company’s business-critical applications, mainframes have become key to Aliant’s business. Aliant’s IT administrators and decision-makers measured the value of mainframes through the business-service management framework, looking closely at the business services the platform supported and how critical those services were to the bottom line.
After a thorough analysis on applications for customer order processing, billing and finance, and credit and collections, Aliant IT staff validated the wisdom of keeping these critical applications on the strong back of the mainframe.
More and more IT organizations are coming to the same conclusion: When looking at an IT organization through the business service management framework, the best platform option might very well be a mainframe.
[1]Gartner picks top four trends, Trevor Trehame, CRN, May 16, 2006