Whilst the mainframe is generally regarded as old hat, many might be surprised to learn that over half of enterprise applications are still heavily reliant upon it.
Working away at the beating heart of enterprise IT for the past half a century, the majority of CIOs believe that the mainframe will continue to be a key business asset over the next decade.
However, much has changed since the 1960s when mainframes rose to prominence.
As businesses have sought to increase their agility and meet customers on their own terms, virtualisation, cloud and mobile have risen to the forefront at the bleeding edge of enterprise IT.
Businesses are now supporting more customer-facing applications than ever before.
However, the mainframe, and more specifically the legacy applications that sit on it, weren’t designed to interact with this new generation of applications.
The end result is that the old and new worlds of enterprise IT have been forced to collide, giving rise to a number of key challenges that make the task of keeping the business running smoothly and efficiently extremely complex.
IT departments must get their houses in order, or run the risk that an unexpected meltdown in the mainframe could have a serious impact on revenues, productivity and customer satisfaction.
Much needed refurbishment
The new applications being written for the cloud, virtualised environments and mobile devices are fundamentally different from their traditional counterparts on the mainframe.
Not only are they more widely distributed and being developed in more complex, modern programming languages, but the teams behind them are generally unfamiliar with the mainframe applications that underpin much of their functionality.
As a result, modern developers often unintentionally introduce inefficiencies or even incompatibilities into enterprise IT. For example, mobile apps alone are driving up mainframe computing capacity by an average of 41%.
This could be significantly reduced if modern developers were able to accept that “performance is a defect” and in fact would optimise their code to work more efficiently with the mainframe.
After 50 years, it’s perhaps not surprising that the experienced developers who programmed in the mainframe environment are reaching retirement age.
The newer generations have grown up using more contemporary point-and-click style interfaces, so the green screen environment of the mainframe is practically a foreign language.
As mainframe skills grow increasingly scarce, businesses must modernise the mainframe development environment in order to future-proof their IT. For example, by introducing more user-friendly dashboards with point-and-click functionality and using development and testing tools that provide deeper insight in application complexity, can help modern development teams to get to grips with the mainframe with far greater speed and efficiency.
Evicting the gremlins
The increasingly distributed nature of enterprise applications is also obscuring problem resolution in the event of IT failures.
As IT environments have extended, so too has the length of time it takes to isolate and resolve the cause of any performance issues that develop; exposing businesses to a heightened risk of unhappy customers and increased costs.
However, despite recognising the risks, the majority of businesses still rely on traditional approaches to managing application performance.
Using system measures such as aggregate data or averages to monitor performance is inadequate in the hyper-distributed world of today’s enterprise IT.
Businesses relying on such measures are often unaware of performance issues until users call in to report them; prolonging problem resolution time.
IT teams must have complete visibility into the end-user experience if they are to manage the risks of mainframe inefficiencies and IT outages effectively.
This means having the ability to trace each and every application transaction, from the end-user’s browser, through the distributed tiers and deep into the mainframe. This approach can provide IT departments with visibility into the impact that distributed code has on the mainframe environment.
These insights can be fed back to development teams, enabling them to optimise application code and reduce the mean time to resolution; thereby reducing mainframe costs and improving customer satisfaction.
The majority of enterprises still operate in technology silos, with a distinct lack of communication and cooperation between various teams in the IT department. This means that operations and development teams are often working with incomplete information about performance issues, causing delays in the problem resolution process.
As a result of this, companies are regularly being forced into ‘war-room’ situations, where the finger of blame is directed around the table while IT teams try to guess at the potential cause of the latest systems error. Not only does this extend the time taken to solve problems, it also ties IT teams up, delaying projects and slowing innovation.
The new problems that have arisen as a result of the modern, joined-up application environment need a new, integrated approach to manage them effectively.
Enterprises must foster a mutually collaborative environment between technology teams in order to reduce the time it takes to solve problems and increase productivity.
In order for this to happen, they must be equipped with the tools that enable them to perform detailed diagnostics on performance issues to quickly and accurately pinpoint the cause.
These tools must equally be integrated across the business, enabling operations and development teams, in multiple regions, and in various time zones around the world to work together to resolve IT performance issues and increase the speed at which the business is able to innovate.
The mainframe is still going strong as it passes its 50th anniversary, and businesses are showing no signs of reducing their reliance on it.
As time goes on and new technologies continue their march into every aspect of business operations, IT complexity and the pressure it places upon the mainframe are only going to increase.
As such, companies must stand up and tackle these issues now so they can remain relevant in a hyper-competitive business landscape.
Sourced from Ian Clarke, mainframe solutions director at Compuware