In this challenging economy, many IT organizations are putting even greater focus on identifying how to best meet business objectives with fewer people and reduced IT budgets. They are discovering how the mainframe can help them deal with these challenges.
Consider the case of a large international bank. The back-office customer account system must journal literally millions of transactions that come in from teller terminals, ATMs, electronic transfers, and online banking web sites. Highly valued, critical business applications cannot afford to have lost transactions, data errors, and other problems, which can be due to a variety of failures. The mainframe is the most resilient, fault-tolerant platform available for keeping the critical applications up and running in order to avoid disrupting the business.
The pressure to cut costs is more urgent then ever before and that has prompted some IT managers to consider moving mature and stable applications and databases off mainframes and onto smaller machines. Yet in many enterprises that have attempted the migration, the results are not encouraging. Many have abandoned the projects after long and costly efforts because they simply could not get them to work. Others have achieved the migration, but performance or reliability has suffered.
Balance Costs and Value
In response to the pressure to reduce costs, you may have set out to "optimize" your data center. This typically involves looking at the applications and databases on each server and determining whether they can be deployed more cost-effectively by consolidating them on a lower-cost platform or moving them to other types of platforms altogether. When you make re-platforming decisions, you must take into account the cost of operating the applications and databases on both the source and the potential target platforms. But cost is only part of the equation. You also have to balance cost against the value of the services that the applications and databases deliver to the business: the goal is to deliver the highest value at the lowest cost.
For example, to determine the business value of an application, you must know which business services and business processes that resource supports. You also need to know the business criticality of those services as measured by the impact that performance degradation or temporary loss of those services would have on the business. A server that supports online stock trading, for example, is of higher value to a brokerage business than a server that supports online access to employee benefits information.
The business value of a resource is underpinned by its inherent capacity, availability, and performance. For example, the mainframe's ability to handle extremely large databases and huge transaction volumes-and do it with unmatched reliability, availability, and serviceability-is crucial in many business-critical applications.
Mainframe specialty processors are great candidates for delivering cost savings. Offloading work to zIIPs can result in huge savings in the total cost per MIPS (as much as $7,000/MIPS), and free up general-purpose engines for the business services work that can't run on the specialty engines.
Monitoring products for z/OS should be able to execute over half their work on a zIIP engine, and the cost savings can be significant. Another way to save is with capacity monitoring products that can model popular specialty processors, enabling you to test-drive moving some of your application work to specialty engines.
Increase Your Reliance Upon the Solid Capabilities of the Mainframe
The mainframe has several unique strengths that make it the right platform for certain jobs, particularly in these tough economic times. These strengths include unmatched capacity; the highest level of reliability, availability, and services; unparalleled business continuity; and resilience. Its exceptional processing and storage capacity enables the mainframe to handle massive databases and huge transaction volumes. Software can be installed and updated without rebooting or otherwise interrupting operation. Mainframes also have extremely high fault tolerance and highly advanced backup and recovery capabilities that permit these platforms to continue operation in the event of a failure or catastrophe in the data center-without missing a single transaction. The software operating environment is also highly resilient.
Create a More Cost-Effective Data Center
What actions can you take to run your data center more cost effectively? First, evaluate the platform assignment using a cost/value analysis and, where necessary, migrate applications and databases to other, more cost-effective platforms. Then, you should optimize your management of the data center to derive maximum business value from applications, databases, and other resources. Business Service Management (BSM) is a comprehensive approach and uniform platform for running IT based on business needs. The solutions should be based on best-practice IT processes, such as those outlined in the IT Infrastructure Library (ITIL) guidelines.
Follow these steps to create a more cost-effective data center:
- Establish a baseline. Determine what resources you have in the data center, their current workload profiles, and their total costs of ownership.
- Assess the business value of your resources. Determine the business services supported by them and assess their value in delivering services.
- Consolidate applications with similar business requirements and deploy them to the most value-effective platforms
Look for solutions that encompass the mainframe as well as other platforms in an integrated fashion. Well-architected BSM solutions can help you right-platform by providing three major capabilities. These include comprehensive monitoring, mapping the relationships among the IT resources and the business services and processes that they support, as well as sophisticated analysis and modeling.
By leveraging solutions to right-platform and manage your data center, you can maximize the value of your mainframes. This process will also help you drive up the value of the other resources in your data center. The results are compelling and include improved service quality, increased staff productivity, faster problem resolution, reduced service desk workload, and other business benefits. These advantages will help your organization meet existing business objectives and you will be better prepared for handling increased demands when the economy turns around.