Mainframe Performance: The Role of the Mainframe in the Application Delivery Chain

Steve Tack

The mainframe has long been established as a secure, scalable and reliable platform for high-value data and high-volume transactions such as enterprise resource planning, online order-taking and financial transaction processing. Today, IT teams continue to rely on the unmatched stability and consistency of mainframe data stores to support business-critical applications. But they're now extracting even more value by 're-incarnating' their mainframes-directly integrating them as part of modern architectures, for example, to provide the rock-solid back-end engines needed for customer-facing, revenue-generating applications. 

Because of the silo-centric nature of a typical IT organization, there's a tendency to overlook the mainframe when it comes to optimizing application performance, or the speed, reliability, security and integrity as experienced by end users. This article will explore the ongoing relevance of the mainframe in application performance management initiatives. Specifically, we will consider application performance as a business issue, the evolution of the mainframe as an engine for both customer-facing and corporate productivity applications, and how even subtle mainframe optimizations can drive end-user application performance improvements that bolster business performance and the bottom line.
Application Performance as a Business Issue

In today's highly interconnected world, application performance is without a doubt a business issue, and applications that perform well-are quick to load and readily accessible to a global base of end users, for example-can drive significant efficiencies and competitive edge. Conversely, applications that don't perform well can severely hamper a business. Consider the example of a large North American fresh food manufacturer that had been experiencing major performance issues in several of its mission-critical logistics applications, including transportation, product tracking and traceability from its farms to its distribution centers. For years, this company had outages in several of these mission-critical applications, and this instability greatly hindered its promise to deliver fresh products. 

By leveraging performance monitoring tools, the organization has increased visibility into the performance of these applications, and is now able to quickly identify root-cause issues that would have otherwise gone unresolved. Today, it is faster at fixing application performance problems and more responsive to application end users. The company also utilizes IT resources more efficiently and has reduced expensive help desk calls by 25-35 percent. 

When it comes to customer-facing Web applications-online order-taking applications, for example-applications that are slow to load, or have periods of unavailability or inconsistent performance can negatively impact end users' experiences. When end users' satisfaction with your application is low, this reduces the likelihood that they will continue to spend time on your site and/or actually go through with a purchase. A recent study analyzing millions of page views on websites around the world found that conversion rates increase 74 percent when page load times improve from eight to two seconds. Another study found that page abandonment rates increase steeply as page load times increase. Similarly, businesses with strong website and application performance report significantly reduced costs in terms of resources and time required to fix and resolve problems, as well as inquiries into costly customer support channels like call centers. Improving website and application performance can increase revenues while lowering costs, a win-win for any business.

The Changing Role of the Mainframe: Supporting Customer-Facing Applications

Throughout the years, mainframe systems have maintained their unique position in supporting some of the most throughput-intensive and business-critical enterprise applications. Mainframes today continue to touch the majority of transactions worldwide, with 72 percent of the world's financial transactions being processed on mainframes, according to Independent Assessment, an industry authority on mainframe computing in the 21st century. Health care providers, government agencies, telecom companies, manufacturers, shipping companies and other industries all rely on the speed, reliability, security and integrity of the mainframe.   

Mainframe applications have changed substantially in the past couple of decades, as they leverage other technology improvements such as the Internet and high-speed global communications. Today's mainframe application paths are increasingly extending beyond the firewall and out to the Web, where they become customer-facing. Banking at home, buying a retailers' product, paying your credit card provider online and making travel reservations are just a few examples of modern applications powered by and expanding the use of mainframe applications.  

Then, there's the explosion of mobile Web-enabled devices. According to Gartner, mobile phones will overtake PCs as the most common Web access devices worldwide by 2013. The massive increase in the variety of activities end users are conducting online-combined with the proliferation of Web-enabled mobile devices-means that the mainframe is being 'touched' way more often than in the past. A high-performance mainframe becomes more essential to employee productivity, partner relations, customer satisfaction, brand, and ultimately the bottom line than ever before.

Mainframes often are, however, just one of a series of elements supporting an overall business service, such as online banking. Finding opportunities to optimize the mainframe-or any element, for that matter-must begin by understanding the true end-user experience. This is especially true for modern applications, where so many elements both within and beyond the firewall (starting with mainframes and then extending all the way out across the Internet to end users' browsers) can impact performance. This is known as the application delivery chain.

Within IT organizations, there's a tendency for various contributors to consider only the health and performance of their own area of responsibility-an individual application or system, for example-without an eye on the health of the overall service being provided. Understanding performance across the entire application delivery chain-and from the true end-user perspective-is the only way to determine if performance for the overall service is sub-par, and if opportunities for optimization exist.

Subtle Mainframe Optimizations Can Make a Big Difference

When it comes to optimizing application performance, you can't overlook the mainframe. Because of the high volume of mainframe transactions, subtle optimizations can exert an enormously positive, measurable impact on the end-user experience.

For example, a leading financial services institution in the U.S. had more than a million online visitors to its retail online banking website on any given day. Performance monitoring tools were surveying these transactions and detected a significant response time problem: The typical response time of a page load was between two and three seconds, and this was rising to an average of 19 seconds for all site visitors. Advanced diagnostics helped the firm isolate the source of the issue to a particular DB2 region on the mainframe. All of the application servers that were calling this region were affected.

By comparing the current performance levels to a historical baseline, the firm saw that response time in DB2 changed from three milliseconds on average to five milliseconds. While this may seem like a trivial time increase, it caused the 3.32 million transactions conducted during this timeframe to slow way down, all the way out at the end users' browsers (in fact, some sessions were even timing out). After identifying, isolating and addressing the source of the problem, database and mainframe administrators were able to bring response time levels back to normal. Given the sheer number of transactions affected, the impact on customer satisfaction and the overall business was huge.

Conclusion:  Mainframe Optimizations Are a Cornerstone of Modern-Day Application Performance Management

Businesses and their IT teams can no longer afford to take a silo-centric, non end-user-focused approach to application performance management. Rather, they need to proactively manage application performance across the entire application delivery chain, beginning with a true view of the end-user experience and exploiting optimization opportunities that deliver the most impact. Due to the tremendous potential gains, IT teams would be remiss to not include the mainframe as part of their application performance management approach. As the mainframe continues to be an essential foundation for modern applications, this will be the key for businesses looking to make the most of their existing mainframe investments and overall IT resources.

Add Comment      Leave a comment on this blog post
Sep 21, 2011 3:09 PM J Ukelson J Ukelson  says:
Steve, I couldn't agree with you more. Mainframes are the most highly virtualized, transaction oriented platform around. Performance is crucial since mainframe still handle about 90% of mission critical business applications in large enterprises. Behavioral analysis is a way to get even more from existing performance monitors - taking the data from those monitors and turning it into actionable knowledge. I've written more about behavioral analytics for the mainframe at: Reply
Oct 14, 2011 8:10 AM Marcel den Hartog Marcel den Hartog  says:
Steve, very good read. The impact of people using mobile devices and the Web to interact with applications was very clear in the past days when BlackBerry's in EMEA went down. People where desperate, complained and it looked like civilization came to a halt. The connection between Mainframes and anything "web" or "mobile" is getting stronger almost daily. Maybe read here for a bit more on that: The problem with managing performance of an application going through a chain of different IT components is the "siloed" approach. Many of the performance tools are platform specific and almost all of them show "The Mainframe" as black hole. Transactions go in, and come out after a while and this lends itself to a lot of fingerpointing and blaming: the "not my problem" attitude unfortunately still exists.... But recently, vendors have seen the light and integrated their MF performance tools with their Distributed brethrens... Offering those who manage an single picture of the transaction, and the possibility to exactly pinpoint where the real problem occurs. Since things are never what they seem in the performance management space, only an integrated approach can guarantee that end-user facing, multi platform applications are as "snappy" as users expect them to be... For those interested in an example: Reply
Mar 4, 2012 7:03 AM happy holi images happy holi images  says:
That's why most of us get very quickly sick and exhausted of seeing the same old image on our desktop's qualifications and we try to modify it regularly. Since we used our computer systems almost every day of our life, it is indeed challenging not to get tired while seeing the same qualifications on our pc. Reply

Post a comment





(Maximum characters: 1200). You have 1200 characters left.



Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.